WO2022027748A1

WO2022027748A1 - Passenger recognition method, passenger recognition apparatus, electronic device and readable storage medium

Info

Publication number: WO2022027748A1
Application number: PCT/CN2020/111433
Authority: WO
Inventors: 赵娟娟; 张刘涛; 须成忠; 叶可江
Original assignee: 中国科学院深圳先进技术研究院
Priority date: 2020-08-04
Filing date: 2020-08-26
Publication date: 2022-02-10
Also published as: CN114090704A; CN114090704B

Abstract

A public transport passenger recognition method, comprising: on the basis of a first passenger travel data source and a second passenger travel data source, respectively acquiring a first travel trajectory that comprises a plurality of first logical travels, and a second travel trajectory that comprises a plurality of second logical travels (S202); matching each first logical travel with each second logical travel, so as to classify the first logical travels, which match the second logical travels, into a matching logical travel set (S204); acquiring an overlapping travel time period of a travel time period of each first logical travel in the matching logical travel set and a travel time period of each second logical travel, so as to obtain an overlap duration corresponding to each overlapping travel time period (S206); performing attenuation processing on each overlap duration to obtain each overlap duration which is subjected to the attenuation processing (S208); calculating the trajectory similarities between the first travel trajectories and the second travel trajectories according to the overlap durations which are subjected to the attenuation processing (S210); and recognizing a first passenger and a second passenger according to the trajectory similarities (S212). Further provided is a public transport passenger recognition apparatus. The accuracy of recognizing the same passenger under different data sources is improved.

Description

乘客的识别方法和识别装置、电子设备和可读存储介质Passenger identification method and identification device, electronic device and readable storage medium

技术领域technical field

本发明属于时空数据处理和公共交通技术领域，具体地讲，涉及一种公共交通乘客的识别方法、公共交通乘客的识别装置、电子设备以及机器可读存储介质。The invention belongs to the technical field of spatiotemporal data processing and public transportation, and in particular, relates to a method for identifying public transportation passengers, an identification device for public transportation passengers, electronic equipment and a machine-readable storage medium.

背景技术Background technique

随着感知技术和网络传输技术的发展，我们可以获取到越来越多的公共交通(诸如地铁、公交车等)乘客出行的数据，例如通过自动收费***(Auto Fare Collection，AFC)收集到的AFC数据(诸如票务数据等)，通过公共交通站安装的移动终端检测(Access Point，AP)设备收集到的携带移动终端的乘客的AP数据(诸如位置数据等)。这些数据的获取为我们分析乘客，特别是个体乘客的出行提供了新的分析思路，例如AFC数据记录了乘客的进出站数据，而AP数据记录了乘客在行程中的部分位置数据，如果将这两种数据相关联，可以获取到乘客完整的出行信息。然而，在实际中，获取不同数据的数据源是彼此相互隔离的，并且同一乘客在不同数据源中所拥有的身份识别信息也是不同的，因此如何将不同数据源中相同乘客相关联是构建乘客详细出行信息的前提。With the development of perception technology and network transmission technology, we can obtain more and more public transportation (such as subway, bus, etc.) passenger travel data, such as the collection of automatic toll collection system (Auto Fare Collection, AFC). AFC data (such as ticketing data, etc.), AP data (such as location data, etc.) of passengers carrying mobile terminals collected by mobile terminal detection (Access Point, AP) equipment installed at public transportation stations. The acquisition of these data provides us with new analysis ideas for analyzing the travel of passengers, especially individual passengers. For example, the AFC data records the passenger's entry and exit data, and the AP data records part of the passenger's position data during the trip. The two kinds of data are associated, and the complete travel information of passengers can be obtained. However, in practice, the data sources for obtaining different data are isolated from each other, and the identification information of the same passenger in different data sources is also different. Therefore, how to associate the same passenger in different data sources is to construct a passenger Prerequisites for detailed travel information.

不同数据源之间的时空轨迹(即乘客的出行轨迹)的关联匹配可以归结为两个时空轨迹的相似性问题，现有的方法主要是对路径不确定的道路交通的时空轨迹的相似性进行比较，即站点之间的比较，然而这样是无法在不同乘客之间存在大量时空轨迹重叠的情况下，实现不同数据源下同一乘客的识别匹配的。图1是现有技术的时空轨迹相似性的比较方法的示意图。The correlation matching of spatiotemporal trajectories (that is, the travel trajectories of passengers) between different data sources can be attributed to the similarity of two spatiotemporal trajectories. Comparison, that is, comparison between stations, however, it is impossible to realize the identification and matching of the same passenger under different data sources when there is a large amount of space-time trajectory overlap between different passengers. FIG. 1 is a schematic diagram of a comparison method of spatiotemporal trajectory similarity in the prior art.

参照图1，可以采集到乘客A和乘客B的以下信息：Referring to Figure 1, the following information of passenger A and passenger B can be collected:

基于AFC***，获取到乘客A经常在一天的固定出行时间段从站点s ₁前往站点s ₅，并且乘客A在出行途中会经过站点s ₂～s ₄，我们将这类出行称作规律出行，如图1中的(a)图所示。然而，在乘客A的一次偶尔出行中，乘客A从站点s ₅前往站点s ₈，我们将这类出行称作随机出行。在图1中的(b)图中，在一天的相同的固定出行时间段，另一位乘客B也经常从站点s ₁前往站点s ₅。 Based on the AFC system, it is obtained that passenger A often travels from station s ₁ to station s ₅ in a fixed travel period of the day, and passenger A will pass through stations s ₂ to s ₄ on the way to travel. We call this type of travel regular travel. As shown in (a) of Figure 1. However, in an occasional trip of passenger A, from station _s5 to station _s8 , we call this type of trip a random trip. In (b) of FIG. 1 , another passenger B also frequently travels from station s ₁ to station s ₅ during the same fixed travel period of the day.

基于AP***，获取到乘客A和乘客B的出行轨迹如图1中的(c)图和(d)图所示，乘客A的出行包括：规律出行(站点s ₂～s ₅、站点s ₁～s ₃)和随机出行(站点s ₅～s ₆)；乘客B的出行包括：规律出行(站点s ₂～s ₅、站点s ₁～s ₃、站点s ₂～s ₄)。 Based on the AP system, the travel trajectories of passenger A and passenger B are obtained as shown in (c) and (d) in Figure 1. Passenger A's travel includes: regular travel (stations s ₂ to s ₅ , station s _{1 )} . ～s ₃ ) and random trips (stops s ₅ ～ s ₆ ); the trips of passenger B include: regular trips (stops s ₂ ～s ₅ , s ₁ ～s ₃ , and s ₂ ～s ₄ ).

如果基于现有的时空轨迹的匹配方法，即进行站点与站点之间的匹配，会获取到错误的匹配，即得到的匹配结果是(a)图中的时空轨迹和(d)图中的时空轨迹相匹配，从而得到乘客A和乘客B相匹配，这是由于不同乘客之间的重叠的时空轨迹(意味着站点也重叠)所导致的。If the matching method based on the existing spatio-temporal trajectories, that is, matching between sites, will obtain an incorrect match, that is, the matching results obtained are (a) the spatio-temporal trajectories in the graph and (d) the spatio-temporal trajectories in the graph. The trajectories are matched, resulting in passenger A and passenger B matching due to overlapping spatiotemporal trajectories between different passengers (meaning that the stations also overlap).

发明内容SUMMARY OF THE INVENTION

为了解决上述现有技术存在的问题，本发明提供了一种能够提高不同数据源下同一乘客的识别匹配准确率的公共交通乘客的识别方法和公共交通乘客的识别装置。In order to solve the above-mentioned problems in the prior art, the present invention provides a public transport passenger identification method and a public transport passenger identification device which can improve the identification and matching accuracy of the same passenger under different data sources.

根据本发明的实施例的一方面提供的公共交通乘客的识别方法，其包括：基于第一乘客出行数据源获取包括多个第一逻辑出行的第一出行轨迹，以及基于第二乘客出行数据源获取包括多个第二逻辑出行的第二出行轨迹，所述第一逻辑出行和所述第二逻辑出行分别包括构成一次逻辑出行的出行时间段；对所述第一逻辑出行和所述第二逻辑出行进行匹配，以将与所述第二逻辑出行匹配的所述第一逻辑出行归类于匹配逻辑出行集；获取所述匹配逻辑出行集中的所述第一逻辑出行的出行时间段和其匹配的所述第二逻辑出行的出行时间段重叠的出行时间段，以得到重叠的出行时间段所对应的重叠时长；对所述重叠时长进行衰减处理，以得到衰减处理后的重叠时长；根据所述衰减处理后的重叠时长计算出所述第一出行轨迹和所述第二出行轨迹的轨迹相似度；根据所述轨迹相似度来对第一乘客和第二乘客进行识别。According to an aspect of an embodiment of the present invention, a method for identifying public transport passengers is provided, which includes: acquiring a first travel trajectory including a plurality of first logical trips based on a first passenger travel data source, and based on a second passenger travel data source Acquire a second travel trajectory including a plurality of second logical trips, the first logical trip and the second logical trip respectively include travel time periods constituting one logical trip; Matching the logical trips to classify the first logical trips matched with the second logical trips into a matching logical trip set; obtaining the trip time period and the corresponding logical trips of the first logical trips in the matching logical trip set matching the travel time period of the second logical travel overlapping the travel time period, to obtain the overlapping time period corresponding to the overlapping travel time period; performing attenuation processing on the overlapping time period to obtain the overlapping time period after attenuation processing; according to The track similarity of the first travel track and the second travel track is calculated from the overlapping duration after the attenuation process; the first passenger and the second passenger are identified according to the track similarity.

可选地，在上述一方面的一个示例中，所述对所述重叠时长进行衰减处理，以得到衰减处理后的重叠时长，具体包括：对所述匹配逻辑出行集中符合预设条件的所述第一逻辑出行所对应的所述重叠时长进行衰减，以得到衰减重叠时长；将所述衰减重叠时长和所述匹配逻辑出行集中不符合所述预设条件的所述第一逻辑出行所对应的所述重叠时长确定为所述衰减处理后的重叠时长；其中，所述预设条件包括：在所述匹配逻辑出行集中的一个出行时间段中，对应该出行时间段的所有所述第一逻辑出行的出行时长之和大于预设时长。Optionally, in an example of the above aspect, the performing attenuation processing on the overlapping durations to obtain the overlapping durations after attenuation processing specifically includes: performing an attenuation process on the overlapping durations that meet the preset conditions in the matching logical line set. The overlapping duration corresponding to the first logical trip is attenuated to obtain the attenuation overlapping duration; the attenuation overlapping duration and the matching logical trip set corresponding to the first logical trip that does not meet the preset condition are combined. The overlapping duration is determined as the overlapping duration after the attenuation process; wherein, the preset condition includes: in a travel time period in the matching logical travel set, all the first logic corresponding to the travel time period The sum of the travel duration of the trip is greater than the preset duration.

可选地，在上述一方面的一个示例中，所述对所述匹配逻辑出行集中符合预设条件的所述第一逻辑出行所对应的所述重叠时长进行衰减，以得到衰减重叠时长，具体包括：基于下面的式子计算得到所述衰减重叠时长，Optionally, in an example of the above aspect, the overlapping duration corresponding to the first logical trip that meets the preset condition in the matching logical trip set is attenuated to obtain the attenuation overlapping duration, specifically: Including: calculating the decay overlap duration based on the following formula,

其中，i表示所述匹配逻辑出行集中的符合所述预设条件的第i个出行时间段，c _i表示所述第i个出行时间段中的所述第一逻辑出行的总数量，j表示所述第i个出行时间段中的第j个第一逻辑出行，γ表示衰减贡献率，

表示所述第j个第一逻辑出行所对应的重叠时长，

表示所述第j个第一逻辑出行所对应的衰减重叠时长。 Wherein, i represents the i-th travel time period that meets the preset condition in the matching logical travel set, c _i represents the total number of the first logical trips in the i-th travel time period, and j represents the jth first logical trip in the ith trip time period, γ represents the decay contribution rate,

represents the overlapping duration corresponding to the jth first logical trip,

Indicates the decay overlap duration corresponding to the jth first logical row.

可选地，在上述一方面的一个示例中，所述第一逻辑出行和所述第二逻辑出行均还包括构成一次逻辑出行的出行最短时长；其中，所述根据所述衰减处理后的重叠时长计算出所述第一出行轨迹和所述第二出行轨迹的轨迹相似度，具体包括：根据所述衰减处理后的重叠时长、所述多个第一逻辑出行的出行最短时长以及与所述多个第一逻辑出行都不匹配的所述第二逻辑出行的出行最短时长，计算出所述第一出行轨迹和所述第二出行轨迹的相似度。Optionally, in an example of the above aspect, both the first logical trip and the second logical trip further include the shortest trip duration that constitutes a logical trip; wherein, the overlap processed according to the attenuation The trajectory similarity between the first travel trajectory and the second travel trajectory is calculated by the duration, which specifically includes: according to the overlapping duration after the attenuation process, the shortest travel duration of the plurality of first logical trips, and the The shortest travel duration of the second logical trip that does not match the multiple first logical trips, and the similarity between the first trip trajectory and the second trip trajectory is calculated.

可选地，在上述一方面的一个示例中，所述第一逻辑出行和所述第二逻辑出行均还包括构成一次逻辑出行的出行最短时长；其中，所述根据所述衰减处理后的重叠时长计算出所述第一出行轨迹和所述第二出行轨迹的轨迹相似度，具体包括：向所述匹配逻辑出行集中的所述第一逻辑出行的出行最短时长、与所述多个第二逻辑出行都不匹配的所述第一逻辑出行的出行最短时长、与所述多个第一逻辑出行都不匹配的所述第二逻辑出行的出行最短时长分别赋予不同的权重系数；根据所述衰减处理后的重叠时长以及被分别赋予不同权重系数的所述匹配逻辑出行集中的所述第一逻辑出行的出行最短时长、与所述多个第二逻辑出行都不匹配的所述第一逻辑出行的出行最短时长、与所述多个第一逻辑出行都不匹配的所述第二逻辑出行的出行最短时长，计算出所述第一出行轨迹和所述第二出行轨迹的相似度。Optionally, in an example of the above aspect, both the first logical trip and the second logical trip further include the shortest trip duration that constitutes a logical trip; wherein, the overlap processed according to the attenuation The trajectory similarity between the first travel trajectory and the second travel trajectory is calculated by the duration, which specifically includes: the shortest travel duration of the first logical trip in the matching logical trip set, and the travel duration of the first logical trip in the matching logical trip set. The shortest trip duration of the first logical trip that does not match the logical trips and the shortest trip duration of the second logical trips that do not match the multiple first logical trips are given different weighting coefficients respectively; according to the The overlapping duration after attenuation processing and the shortest trip duration of the first logical trip in the matching logical trip set assigned different weight coefficients, and the first logical trip that does not match the plurality of second logical trips The shortest trip duration of the trip and the shortest trip duration of the second logical trip that do not match the multiple first logical trips, and the similarity between the first trip trajectory and the second trip trajectory is calculated.

可选地，在上述一方面的一个示例中，所述第一逻辑出行和所述第二逻辑出行均还包括构成一次逻辑出行的进站站点、出站站点、进站时间和出站时间；其中，所述对所述第一逻辑出行和所述第二逻辑出行进行匹配，以将与所述第二逻辑出行匹配的所述第一逻辑出行归类于匹配逻辑出行集，具体包括：利用所述第一逻辑出行的进站时间减去所述第二逻辑出行的进站时间以得到进站时长，以及利用所述第二逻辑出行的出站时间减去所述第一逻辑出行的出站时间以得到出站时长；获取从所述第一逻辑出行的进站站点到所述第二逻辑出行的进站站点之间的最短进站时长，以及获取从所述第二逻辑出行的出站站点到所述第一逻辑出行的出站站点之间的最短出站时长；若所述进站时长和所述最短进站时长之和的绝对值小于一预设时长阈值，且所述出站时长和所述最短出站时长之和的绝对值小于所述预设时长阈值，则确定所述第一逻辑出行和所述第二逻辑出行匹配，并将与所述第二逻辑出行匹配的所述第一逻辑出行归类于匹配逻辑出行集。Optionally, in an example of the above aspect, both the first logical trip and the second logical trip further include an inbound site, an outbound site, an inbound time, and an outbound time that constitute a logical trip; The matching of the first logical trip and the second logical trip to classify the first logical trip matching the second logical trip into a matching logical trip set specifically includes: using The inbound time of the first logical trip is subtracted from the inbound time of the second logical trip to obtain the inbound duration, and the outbound time of the first logical trip is subtracted from the outbound time of the second logical trip. station time to get the outbound duration; get the shortest inbound duration from the inbound station of the first logical trip to the inbound station of the second logical trip, and get the outbound time from the second logical trip The shortest outbound duration between the station site and the outbound site of the first logical trip; if the absolute value of the sum of the inbound duration and the shortest inbound duration is less than a preset duration threshold, and the outbound duration The absolute value of the sum of the station duration and the shortest outbound duration is less than the preset duration threshold, then it is determined that the first logical trip matches the second logical trip, and the one that matches the second logical trip is determined. The first logical trip is classified into a set of matching logical trips.

根据本发明的实施例的另一方面提供的公共交通乘客的识别装置，其包括：出行轨迹获取模块，被配置为基于第一乘客出行数据源获取包括多个第一逻辑出行的第一出行轨迹，且基于第二乘客出行数据源获取包括多个第二逻辑出行的第二出行轨迹；其中，所述第一逻辑出行和所述第二逻辑出行分别包括构成一次逻辑出行的出行时间段；逻辑出行匹配模块，被配置为对所述第一逻辑出行和所述第二逻辑出行进行匹配，以将与所述第二逻辑出行匹配的所述第一逻辑出行归类于匹配逻辑出行集；重叠时长获取模块，被配置为获取所述匹配逻辑出行集中的所述第一逻辑出行的出行时间段和其匹配的所述第二逻辑出行的出行时间段重叠的出行时间段，以得到重叠的出行时间段所对应的重叠时长；重叠时长衰减模块，被配置为对所述重叠时长进行衰减处理，以得到衰减处理后的重叠时长；轨迹相似度计算模块，被配置为根据所述衰减处理后的重叠时长计算出所述第一出行轨迹和所述第二出行轨迹的轨迹相似度；乘客识别模块，被配置为根据所述轨迹相似度对第一乘客和第二乘客进行识别。According to another aspect of an embodiment of the present invention, a public transportation passenger identification device is provided, comprising: a travel trajectory acquisition module configured to acquire a first travel trajectory including a plurality of first logical travels based on a first passenger travel data source , and obtain a second travel trajectory including a plurality of second logical travels based on the second passenger travel data source; wherein, the first logical travel and the second logical travel respectively include travel time periods that constitute a logical travel; the logic a trip matching module configured to match the first logical trip and the second logical trip to classify the first logical trip matching the second logical trip into a set of matching logical trips; overlapping A duration acquisition module configured to acquire the travel time period in which the travel time period of the first logical trip in the matching logical trip set and the travel time period of the matching second logical trip overlap, so as to obtain the overlapping trip the overlapping duration corresponding to the time segment; the overlapping duration attenuation module is configured to perform attenuation processing on the overlapping duration to obtain the overlapping duration after attenuation processing; the trajectory similarity calculation module is configured to perform attenuation processing according to the attenuation processing. The overlapping duration calculates the trajectory similarity of the first travel trajectory and the second travel trajectory; the passenger identification module is configured to identify the first passenger and the second passenger according to the trajectory similarity.

可选地，在上述另一方面的一个示例中，所述重叠时长衰减模块包括：时长衰减单元，被配置为对所述匹配逻辑出行集中符合预设条件的所述第一逻辑出行所对应的所述重叠时长进行衰减，以得到衰减重叠时长；时长确定单元，被配置为将所述衰减重叠时长和所述匹配逻辑出行集中不符合所述预设条件的所述第一逻辑出行所对应的所述重叠时长确定为所述衰减处理后的重叠时长；其中，所述预设条件包括：在所述匹配逻辑出行集中的一个出行时间段中，对应该出行时间段的所有所述第一逻辑出行的出行时长之和大于预设时长。Optionally, in an example of the above aspect, the overlapping duration attenuation module includes: a duration attenuation unit configured to The overlapping duration is attenuated to obtain an attenuated overlapping duration; a duration determining unit is configured to compare the attenuated overlapping duration and the first logical trip in the matching logical trip set corresponding to the first logical trip that does not meet the preset condition. The overlapping duration is determined as the overlapping duration after the attenuation process; wherein, the preset condition includes: in a travel time period in the matching logical travel set, all the first logic corresponding to the travel time period The sum of the travel duration of the trip is greater than the preset duration.

可选地，在上述另一方面的一个示例中，所述时长衰减单元进一步被配置为利用下面的式子计算得到所述衰减重叠时长，Optionally, in an example of the above aspect, the duration attenuation unit is further configured to obtain the attenuation overlap duration by calculating the following formula:

表示所述第j个第一逻辑出行所对应的重叠时长，

可选地，在上述另一方面的一个示例中，所述第一逻辑出行和所述第二逻辑出行均还包括构成一次逻辑出行的出行最短时长；所述轨迹相似度计算模块进一步被配置为：根据所述衰减处理后的重叠时长、所述多个第一逻辑出行的出行最短时长以及与所述多个第一逻辑出行都不匹配的所述第二逻辑出行的出行最短时长，计算出所述第一出行轨迹和所述第二出行轨迹的相似度。Optionally, in an example of the above aspect, both the first logical trip and the second logical trip further include the shortest trip duration that constitutes a logical trip; the trajectory similarity calculation module is further configured to: : Calculated according to the overlapping duration after the attenuation process, the shortest travel duration of the multiple first logical trips, and the shortest trip duration of the second logical trip that does not match the multiple first logical trips The similarity between the first travel trajectory and the second travel trajectory.

可选地，在上述另一方面的一个示例中，所述第一逻辑出行和所述第二逻辑出行均还包括构成一次逻辑出行的出行最短时长；其中，所述轨迹相似度计算模块包括：权重系数赋予单元，被配置为向所述匹配逻辑出行集中的所述第一逻辑出行的出行最短时长、与所述多个第二逻辑出行都不匹配的所述第一逻辑出行的出行最短时长、与所述多个第一逻辑出行都不匹配的所述第二逻辑出行的出行最短时长分别赋予不同的权重系数；相似度计算单元，被配置为根据所述衰减处理后的重叠时长以及被分别赋予不同权重系数的所述匹配逻辑出行集中的所述第一逻辑出行的出行最短时长、与所述多个第二逻辑出行都不匹配的所述第一逻辑出行的出行最短时长、与所述多个第一逻辑出行都不匹配的所述第二逻辑出行的出行最短时长，计算出所述第一出行轨迹和所述第二出行轨迹的轨迹相似度。Optionally, in an example of the above aspect, both the first logical trip and the second logical trip further include the shortest trip duration that constitutes a logical trip; wherein, the trajectory similarity calculation module includes: A weighting coefficient assigning unit configured to assign the shortest travel duration of the first logical trip in the matching logical trip set and the shortest trip duration of the first logical trip that does not match the plurality of second logical trips , and the shortest travel durations of the second logical trips that do not match the multiple first logical trips are respectively assigned different weight coefficients; the similarity calculation unit is configured to The shortest travel duration of the first logical trip in the matching logical trip set with different weight coefficients, the shortest trip duration of the first logical trip that does not match the plurality of second logical trips, and the The shortest travel duration of the second logical trip that does not match the multiple first logical trips is calculated, and the trajectory similarity between the first trip trajectory and the second trip trajectory is calculated.

可选地，在上述另一方面的一个示例中，所述第一逻辑出行和所述第二逻辑出行均还包括构成一次逻辑出行的进站站点、出站站点、进站时间和出站时间；所述逻辑出行匹配模块包括：进出站时长确定单元，被配置为利用所述第一逻辑出行的进站时间减去所述第二逻辑出行的进站时间以得到进站时长，以及利用所述第二逻辑出行的出站时间减去所述第一逻辑出行的出站时间以得到出站时长；最短进出站时长确定单元，被配置为获取从所述第一逻辑出行的进站站点到所述第二逻辑出行的进站站点之间的最短进站时长，以及获取从所述第二逻辑出行的出站站点到所述第一逻辑出行的出站站点之间的最短出站时长；匹配确定单元，被配置为若所述进站时长和所述最短进站时长之和的绝对值小于一预设时长阈值，且所述出站时长和所述最短出站时长之和的绝对值小于所述预设时长阈值，则确定所述第一逻辑出行和所述第二逻辑出行匹配，并将与所述第二逻辑出行匹配的所述第一逻辑出行归类于匹配逻辑出行集。Optionally, in an example of the above aspect, both the first logical trip and the second logical trip further include an inbound station, an outbound station, an inbound time, and an outbound time that constitute a logical trip. ; the logical trip matching module includes: an inbound and outbound duration determination unit configured to use the inbound time of the first logical trip to subtract the inbound time of the second logical trip to obtain the inbound duration, and use all the The outbound time of the second logical trip is subtracted from the outbound time of the first logical trip to obtain the outbound duration; the shortest inbound and outbound duration determination unit is configured to obtain the distance from the inbound station of the first logical trip to the shortest inbound duration between inbound stations of the second logical trip, and obtaining the shortest outbound duration from the outbound station of the second logical trip to the outbound station of the first logical trip; The matching determination unit is configured to, if the absolute value of the sum of the inbound duration and the shortest inbound duration is less than a preset duration threshold, and the absolute value of the sum of the outbound duration and the shortest outbound duration If it is less than the preset duration threshold, it is determined that the first logical trip matches the second logical trip, and the first logical trip matching the second logical trip is classified into a matching logical trip set.

根据本发明的实施例的又一方面提供的电子设备，其包括：至少一个处理器，以及与所述至少一个处理器耦合的存储器，所述存储器存储指令，当所述指令被所述至少一个处理器执行时，使得所述至少一个处理器执行如上所述的公共交通乘客的识别方法。According to yet another aspect of an embodiment of the present invention, an electronic device is provided, comprising: at least one processor, and a memory coupled with the at least one processor, the memory storing instructions, when the instructions are executed by the at least one processor When executed by the processor, the at least one processor is caused to execute the method for identifying a public transportation passenger as described above.

根据本发明的实施例的再一方面提供的机器可读存储介质，其存储有可执行指令，所述指令当被执行时使得所述机器执行如上所述的公共交通乘客的识别方法。Still another aspect of an embodiment of the present invention provides a machine-readable storage medium storing executable instructions that, when executed, cause the machine to perform the above-described method for identifying a public transportation passenger.

有益效果：通过获取相互匹配的第一逻辑出行和第二逻辑出行的重叠时长，并对获取到重叠时长进行衰减，以在计算轨迹相似度的过程中降低重叠时长的权重，从而提高不同数据源下同一乘客的识别匹配准确率。Beneficial effect: By acquiring the overlapping durations of the first logical trip and the second logical trip that match each other, and attenuating the acquired overlapping durations, the weight of the overlapping durations can be reduced in the process of calculating the similarity of the trajectory, thereby improving different data sources. The identification matching accuracy rate of the same passenger.

附图说明Description of drawings

通过结合附图进行的以下描述，本发明的实施例的上述和其它方面、特点和优点将变得更加清楚，附图中：The above and other aspects, features and advantages of embodiments of the present invention will become more apparent from the following description taken in conjunction with the accompanying drawings, in which:

图1是现有技术的时空轨迹相似性的比较方法的示意图；Fig. 1 is the schematic diagram of the comparison method of the spatiotemporal trajectory similarity of the prior art;

图2是根据本发明的实施例的公共交通乘客的识别方法的流程图；2 is a flowchart of a method for identifying a public transportation passenger according to an embodiment of the present invention;

图3A是根据本发明的实施例的公共交通乘客的识别装置的框图；3A is a block diagram of an apparatus for identifying public transportation passengers according to an embodiment of the present invention;

图3B是图3A中的识别装置的逻辑出行匹配模块的一个示例的框图；Figure 3B is a block diagram of an example of a logical row matching module of the identification device in Figure 3A;

图3C是图3A中的识别装置的重叠时长衰减模块的一个示例的框图；3C is a block diagram of an example of an overlap duration decay module of the identification device in FIG. 3A;

图3D是图3A中的识别装置的轨迹相似度计算模块的一个示例的框图；3D is a block diagram of an example of a trajectory similarity calculation module of the identification device in FIG. 3A;

图4是示出了根据本发明的实施例的实现公共交通乘客的识别方法的电子设备的方框图。FIG. 4 is a block diagram illustrating an electronic device implementing a method for identifying a public transportation passenger according to an embodiment of the present invention.

具体实施方式detailed description

以下，将参照附图来详细描述本发明的具体实施例。然而，可以以许多不同的形式来实施本发明，并且本发明不应该被解释为限制于这里阐述的具体实施例。相反，提供这些实施例是为了解释本发明的原理及其实际应用，从而使本领域的其他技术人员能够理解本发明的各种实施例和适合于特定预期应用的各种修改。Hereinafter, specific embodiments of the present invention will be described in detail with reference to the accompanying drawings. The present invention may, however, be embodied in many different forms and should not be construed as limited to the specific embodiments set forth herein. Rather, these embodiments are provided to explain the principles of the invention and its practical application, to thereby enable others skilled in the art to understand the invention for various embodiments and with various modifications as are suited to the particular intended use.

如本文中使用的，术语“包括”及其变型表示开放的术语，含义是“包括但不限于”。术语“基于”、“根据”等表示“至少部分地基于”、“至少部分地根据”。术语“一个实施例”和“一实施例”表示“至少一个实施例”。术语“另一个实施例”表示“至少一个其他实施例”。术语“第一”、“第二”等可以指代不同的或相同的对象。下面可以包括其他的定义，无论是明确的还是隐含的。除非上下文中明确地指明，否则一个术语的定义在整个说明书中是一致的。As used herein, the term "including" and variations thereof represent open-ended terms meaning "including but not limited to". The terms "based on", "depending on" and the like mean "based at least in part on", "based at least in part on". The terms "one embodiment" and "an embodiment" mean "at least one embodiment." The term "another embodiment" means "at least one other embodiment." The terms "first", "second", etc. may refer to different or the same objects. Other definitions, whether explicit or implicit, may be included below. The definition of a term is consistent throughout the specification unless the context clearly dictates otherwise.

如背景技术所述，在公共交通出行中，不同乘客之间存在大量重叠的时空轨迹，在时空轨迹识别过程中重叠的时空轨迹会影响不同数据源下同一乘客的识别匹配准确率。因此，如果在时空轨迹的匹配过程中，能够降低重叠的时空轨迹(即规律出行)的权重，那么将会提高不同数据源下同一乘客的识别匹配准确率。As described in the background art, there are a large number of overlapping spatiotemporal trajectories among different passengers in public transport travel, and the overlapping spatiotemporal trajectories in the process of spatiotemporal trajectory identification will affect the identification and matching accuracy of the same passenger under different data sources. Therefore, if the weight of overlapping spatiotemporal trajectories (that is, regular travel) can be reduced in the process of matching spatio-temporal trajectories, the accuracy of identifying and matching the same passenger under different data sources will be improved.

为了实现上述目的，根据本发明的实施例提供了一种公共交通乘客的识别方法。该识别方法包括：基于第一乘客出行数据源获取包括多个第一逻辑出行的第一出行轨迹，以及基于第二乘客出行数据源获取包括多个第二逻辑出行的第二出行轨迹，第一逻辑出行和第二逻辑出行分别包括构成一次逻辑出行的出行时间段；对第一逻辑出行和第二逻辑出行进行匹配，以将与第二逻辑出行匹配的第一逻辑出行归类于匹配逻辑出行集；获取匹配逻辑出行集中的第一逻辑出行的出行时间段和其匹配的第二逻辑出行的出行时间段重叠的出行时间段，以得到重叠的出行时间段所对应的重叠时长；对重叠时长进行衰减处理，以得到衰减处理后的重叠时长；根据衰减处理后的重叠时长计算出第一出行轨迹和第二出行轨迹的轨迹相似度；根据轨迹相似度来对第一乘客和第二乘客进行识别。当然，这里的识别指的是识别第一乘客和第二乘客是否是同一乘客。In order to achieve the above object, a method for identifying a public transportation passenger is provided according to an embodiment of the present invention. The identification method includes: acquiring a first travel trajectory including a plurality of first logical trips based on a first passenger travel data source, and acquiring a second travel trajectory including a plurality of second logical trips based on a second passenger travel data source, the first The logical trip and the second logical trip respectively include a trip time period that constitutes a logical trip; the first logical trip and the second logical trip are matched to classify the first logical trip matching the second logical trip into the matching logical trip set; obtain the travel time period overlapping the travel time period of the first logical trip in the matching logical travel set and the travel time period of the matching second logical trip, so as to obtain the overlapping time period corresponding to the overlapping travel time period; Perform attenuation processing to obtain the overlapping duration after attenuation processing; calculate the trajectory similarity of the first travel trajectory and the second travel trajectory according to the overlapping duration after attenuation processing; calculate the trajectory similarity between the first passenger and the second passenger according to the trajectory similarity. identify. Of course, the identification here refers to identifying whether the first passenger and the second passenger are the same passenger.

在该识别方法中，通过获取匹配的第一逻辑出行和第二逻辑出行的重叠时长，并对获取到重叠时长进行衰减，以在计算轨迹相似度的过程中降低重叠时长的权重，从而提高不同数据源下同一乘客的识别匹配准确率。In this identification method, the overlapping duration of the matching first logical trip and the second logical trip is obtained, and the obtained overlapping duration is attenuated, so as to reduce the weight of the overlapping duration in the process of calculating the similarity of the trajectory, thereby increasing the difference between the two. The identification matching accuracy rate of the same passenger under the data source.

以下，在详细描述根据本发明的实施例的公共交通乘客的识别方法和识别装置之前，先对用于本发明的各实施例的一些术语概念进行说明。Hereinafter, before describing in detail the identification method and identification device for public transportation passengers according to embodiments of the present invention, some terminology concepts used in various embodiments of the present invention will be explained.

出行数据源：其能够在乘客出行时获取并存储乘客的乘客标识、站点数据、时间数据等的乘客出行数据。Travel data source: It can acquire and store passenger travel data such as passenger identification, station data, time data, etc. of passengers when they travel.

逻辑出行：是指乘客的一次逻辑的乘车过程，即一次乘客从进站站点进站，并从出站站点出站的全过程。在一个示例中，一个逻辑出行可以包括：进站站点、进站时间、出站站点、出站时间等。这里，进站时间和出站时间构成出行时间段，因此，一个逻辑出行还可以包括：出行时间段。此外，出行时间段所对应的时长被设定为出行时长。在一个示例中，乘客在一天的早上8时从进站站点a进站，并在该天的早上9时从出站站点b出站，这样的一次逻辑的乘车过程，构成一个逻辑出行；其中，进站时间为早上8时，出站时间为早上9时，那么出行时间段为早上8时～早上9时，其所对应的出行时长为1个小时。Logical travel: refers to a logical ride process of passengers, that is, the whole process of a passenger entering the station from the inbound station and leaving the station from the outbound station. In one example, a logical trip may include: inbound station, inbound time, outbound station, outbound time, and the like. Here, the inbound time and the outbound time constitute the travel time period, therefore, a logical trip can also include: travel time period. In addition, the time period corresponding to the travel time period is set as the travel time period. In an example, passengers enter the station from the inbound station a at 8:00 in the morning, and leave the station from the outbound station b at 9:00 in the morning of the day, such a logical ride process constitutes a logical trip; Among them, the entry time is 8:00 am and the exit time is 9:00 am, then the travel time period is from 8:00 am to 9:00 am, and the corresponding travel time is 1 hour.

出行最短时长：给定任意两个物理站点，从其中一个物理站点到另一个物理站点所经过的最短路径(也称有效路径)所用的时长。在一个示例中，给定两个物理站点，分别为站点a和站点b，从站点a到站点b，存在两条路径(一般可以通过公共交通路线获知)，其中一条路径是从站点a，途径站点a1、站点a2而到达站点b，而另一条路径是从站点a，途径站点b1而到达站点b，那么从站点a到站点b的出行最短时长是从站点a，途径站点b1而到达站点b所用的时长，当然此处的假设前提是任意两个物理站点之间的途径距离基本相等，这也符合实际公共交通的设置。因此，在一个示例中，一个逻辑出行还可以包括出行最短时长。Shortest travel duration: Given any two physical sites, the duration of the shortest path (also called the effective path) from one physical site to the other. In one example, given two physical stations, station a and station b, from station a to station b, there are two paths (generally known by public transportation routes), one of which is from station a, via Station a1, station a2 to station b, and another route is from station a, through station b1 to station b, then the shortest travel time from station a to station b is from station a, through station b1 to station b The duration used, of course, the premise here is that the route distance between any two physical stations is basically equal, which is also in line with the actual public transportation settings. Thus, in one example, a logical trip may also include a minimum trip duration.

出行轨迹：指由在一段时间内乘客被感知设备(例如AFC***、AP***)采集到的按照时间先后顺序排列的多个时空点(或称检测点)构成的时空点序列。在一个示例中，每个时空点包括安装有感知设备的站点和相应的时间(即感知检测的时间)。Travel trajectory: refers to the spatio-temporal point sequence composed of multiple spatio-temporal points (or detection points) arranged in chronological order and collected by the passengers in a period of time by sensing equipment (such as AFC system, AP system). In one example, each spatiotemporal point includes the site where the sensing device is installed and the corresponding time (ie, the time of sensing detection).

以上是对用于本发明的各实施例的一些术语概念进行了说明，接下来，将结合附图来详细描述根据本发明的实施例的公共交通乘客的识别方法以及公共交通乘客的识别装置。The above describes some terminology concepts used in various embodiments of the present invention. Next, a method for identifying public transport passengers and an apparatus for identifying public transport passengers according to embodiments of the present invention will be described in detail with reference to the accompanying drawings.

图2是根据本发明的实施例的公共交通乘客的识别方法的流程图。FIG. 2 is a flowchart of a method for identifying a public transportation passenger according to an embodiment of the present invention.

参照图2，在框S202中，基于第一乘客出行数据源获取包括多个第一逻辑出行的第一出行轨迹，以及基于第二乘客出行数据源获取包括多个第二逻辑出行的第二出行轨迹。2, in block S202, a first travel trajectory including a plurality of first logical travels is obtained based on a first passenger travel data source, and a second travel including a plurality of second logical travels is obtained based on the second passenger travel data source trajectory.

在根据本发明的实施例中，第一乘客出行数据源能够获取并存储第一乘客在出行时，第一乘客所有的进站站点、进站时间、出站站点、出站时间、乘客标识(即身份标识)等。在一个示例中，第一乘客出行数据源可以例如是AFC数据源(即AFC***)。In the embodiment according to the present invention, the travel data source of the first passenger can acquire and store all the inbound sites, inbound time, outbound sites, outbound time, passenger identification ( identification), etc. In one example, the first passenger travel data source may be, for example, an AFC data source (ie, an AFC system).

第一出行轨迹包括从第一乘客出行数据源获取到的多个进站站点、与各进站站点对应的进站时间、多个出站站点、与各出站站点对应的出站时间。因此，每个第一逻辑出行包括对应的一个进站站点、一个进站时间、一个出站站点、一个出站时间。如上所述，根据进站时间和出站时间可以获得出行时间段及其对应的出行时长，因此每个第一逻辑出行还可以包括对应的一个出行时间段及一个出行时长。还如上所述，根据进站站点和出站站点可以获得最短出行时长，因此每个第一逻辑出行还可以包括对应的一个最短出行时长。The first travel trajectory includes a plurality of inbound sites obtained from the first passenger travel data source, an inbound time corresponding to each inbound site, a plurality of outbound sites, and an outbound time corresponding to each outbound site. Therefore, each first logical trip includes a corresponding one inbound station, one inbound time, one outbound station, and one outbound time. As described above, the travel time period and its corresponding travel duration can be obtained according to the inbound time and the outbound time, so each first logical trip may also include a corresponding travel time period and a corresponding travel duration. As also described above, the shortest travel duration can be obtained according to the inbound station and the outbound station, so each first logical trip may further include a corresponding shortest travel duration.

在根据本发明的实施例中，第二乘客出行数据源能够获取并存储第二乘客出行时第二乘客所有的途径站点、与各途径站点对应的途径时间、乘客标识(即身份标识)等。在一个示例中，第二乘客出行数据源可例如是AP数据源。应当说明的是，所述途径站点是在第二乘客出行时所途径的公共交通站点，当然可以是第二乘客在进站站点或出站站点时被感知检测而获取到的站点。In the embodiment according to the present invention, the travel data source of the second passenger can acquire and store all the route stations of the second passenger when the second passenger travels, the route time corresponding to each route station, and the passenger identification (i.e., identity identification) and the like. In one example, the second passenger travel data source may be, for example, an AP data source. It should be noted that the route station is a public transportation station passed by the second passenger when he travels, and certainly may be a station acquired by the second passenger through sensing and detection when entering or exiting the station.

第二出行轨迹包括从第二乘客出行数据源获取到的多个途径站点以及各个途径站点各自所对应的途径时间。在一个示例中，按照途径时间先后顺序排列所述多个途径站点，如果当前的途径站点p _i与前一个途径站点p _i-1满足划分条件

则将前一个途径站点p _i-1(以及其对应的途径时间)及该途径站点p _i-1之前的所有途径站点(以及各自所对应的途径时间)包括于一个第二逻辑出行中，其中，p _i.t表示途径站点p _i所对应的途径时间，p _i-1.t表示途径站点p _i-1所对应的途径时间，

表示从途径站点p _i到前一个途径站点p _i-1的最短出行时长，δ为可以容忍的额外附加时长。应当说明的是，前一个途径站点p _i-1之前的途径站点指的是前一个途径站点p _i-1之前的没有满足所述划分条件且没有包括于其他的第二逻辑出行中的途径站点。此外，需要说明的是，如果当前的途径站点与前一个途径站点满足所述划分条件，并且与后一个途径站点也满足所述划分条件，则将当前的途径站点删除。也就是说，将按所述划分条件划分后孤立的途径站点删除。在这种情况下，每个第二逻辑出行包括至少两个途径站点。 The second travel track includes a plurality of route stations obtained from the second passenger travel data source and the route time corresponding to each route station. In one example, the plurality of pathway sites are arranged in chronological order of pathways, if the current pathway site p _i and the previous pathway site p _i-1 satisfy the division condition

Then, the previous route site p _i-1 (and its corresponding route time) and all route sites before this route site p _i-1 (and their corresponding route times) are included in a second logical trip, wherein , p i.t represents the pathway time corresponding to the pathway site p _i , p _i _- 1.t represents the pathway time corresponding to the pathway site p _i-1 ,

represents the shortest travel time from the route station _pi to the previous route station pi _-1 , and δ is the additional additional time that can be tolerated. It should be noted that the route station before the previous route station p _i-1 refers to the route station before the previous route station p _i-1 that does not meet the above-mentioned division conditions and is not included in other second logical trips. . In addition, it should be noted that, if the current route station and the previous route station satisfy the division condition, and also satisfy the division condition with the latter route station, the current route station is deleted. That is to say, the isolated pathway sites after being divided according to the dividing conditions are deleted. In this case, each second logical trip includes at least two waypoints.

在一个示例中，将每个第二逻辑出行中按时间先后顺序排序的至少两个途径站点的第一个途径站点(所对应的途径时间最早)设定为属于该第二逻辑出行中的进站站点，而将最后一个途径站点(所对应的途径时间最晚)设定为属于该第二逻辑出行中的出站站点。在这种情况下，第一个途径站点对应的途径时间即为进站时间，而最后一个途径站点所对应的途径时间即为出站时间。如上所述，根据进站时间和出站时间可以获得出行时间段及其对应的出行时长，因此每个第二逻辑出行还可以包括对应的一个出行时间段和一个出行时长。还如上所述，根据进站站点和出站站点可以获得最短出行时长，因此每个第二逻辑出行还可以包括对应的一个最短出行时长。In one example, the first route stop (the corresponding route time is the earliest) of the at least two route stops in chronological order in each second logical trip is set as belonging to the route in the second logical trip. station, and set the last route station (with the latest route time) as the outbound station belonging to the second logical trip. In this case, the route time corresponding to the first route station is the inbound time, and the route time corresponding to the last route station is the outbound time. As described above, the travel time period and its corresponding travel duration can be obtained according to the inbound time and the outbound time, so each second logical trip may also include a corresponding travel time period and a corresponding travel duration. As also described above, the shortest travel duration can be obtained according to the inbound station and the outbound station, so each second logical trip may further include a corresponding shortest travel duration.

在框S204中，对各个第一逻辑出行和各个第二逻辑出行进行匹配，以将与第二逻辑出行匹配的第一逻辑出行归类于匹配逻辑出行集。In block S204, the respective first logical trips and the respective second logical trips are matched to classify the first logical trips matching the second logical trips into a matching logical trip set.

在一个示例中，如果第i个第一逻辑出行和第j个第二逻辑出行满足下面的条件，则认定第i个第一逻辑出行和第j个第二逻辑出行相互匹配。In one example, if the ith first logical trip and the jth second logical trip satisfy the following conditions, the ith first logical trip and the jth second logical trip are deemed to match each other.

需要满足的条件可以是指定的。在一个示例中，需要满足的条件可以包括如下的条件：The conditions that need to be met can be specified. In one example, the conditions to be satisfied may include the following conditions:

首先，利用第i个第一逻辑出行的进站时间

减去第j个第二逻辑出行的进站时间

以得到进站时长，以及利用第j个第二逻辑出行的出站时间

减去第i个第一逻辑出行的出站时间

以得到出站时长。 First, use the inbound time of the i-th first logical trip

Subtract the pit stop time of the jth second logical trip

to get the inbound duration and the outbound time of the trip using the jth second logic

Subtract the outbound time of the ith first logical trip

to get the outbound time.

其次，获取从第i个第一逻辑出行的进站站点

到第j个第二逻辑出行的进站站点

之间的最短进站时长

且获取从第j个第二逻辑出行的出站站点

到第i个第一逻辑出行的出站站点

之间的最短出站时长

Second, get the pit stop from the ith first logical trip

to the pit stop of the jth second logical trip

Minimum pit stop time between

and get the outbound stop from the jth second logical trip

To the outbound stop of the i-th first logical trip

Minimum outbound time between

最后，如果所述进站时长和所述最短进站时长

之和的绝对值小于一预设时长阈值ε'，且所述出站时长和所述最短出站时长

之和的绝对值小于所述预设时长阈值ε'，则认为第i个第一逻辑出行和第j个第二逻辑出行相互匹配，并将与第j个第二逻辑出行匹配的第i个第一逻辑出行归类于匹配逻辑出行集PG。 Finally, if the pit stop duration and the shortest pit stop duration

The absolute value of the sum is less than a preset duration threshold ε', and the outbound duration and the shortest outbound duration

The absolute value of the sum is less than the preset duration threshold ε', then it is considered that the i-th first logical trip and the j-th second logical trip match each other, and the i-th logical trip that matches the j-th second logical trip The first logical trip is classified into the matching logical trip set PG.

换句话讲，在这个示例中，需要满足的条件可以由下面的式子1表达。In other words, in this example, the condition that needs to be satisfied can be expressed by Equation 1 below.

[1]

且

[1]

and

在另一示例中，需要满足的条件除包括上述条件之外，还可以包括如下条件：第i个第一逻辑出行的进站站点

和出站站点

之间的有效路径经过第j个第二逻辑出行的进站站点

和出站站点

In another example, in addition to the above-mentioned conditions, the conditions to be satisfied may also include the following conditions: the inbound station of the i-th first logical trip

and outbound sites

The valid path between passes through the inbound stop of the jth second logical trip

and outbound sites

此外，在根据本发明的实施例中，对各个第一逻辑出行和各个第二逻辑出行进行匹配之后，将与第二出行轨迹的各个第二逻辑出行都不匹配的第一逻辑出行归类于第一未匹配逻辑出行集NPG1，并将与第一出行轨迹的各个第一逻辑出行都不匹配的第二逻辑出行归类于第二未匹配逻辑出行集NPG2。这里，不匹配指的是不满足上述的条件。在一个示例中，如果不满足上面的式子1，则第i个第一逻辑出行和第j个第二逻辑出行不匹配。在另一示例中，如果不满足“第i个第一逻辑出行的进站站点

和出站站点

之间的有效路径经过第j个第二逻辑出行的进站站点

和出站站点”的条件，则第i个第一逻辑出行和第j个第二逻辑出行不匹配。在又一个示例中，如果不满足上面的式子1，并且不满足“第i个第一逻辑出行的进站站点

和出站站点

之间的有效路径经过第j个第二逻辑出行的进站站点

和出站站点”，则第i个第一逻辑出行和第j个第二逻辑出行不匹配。 In addition, in the embodiment according to the present invention, after each first logical trip and each second logical trip are matched, the first logical trip that does not match each second logical trip of the second trip trajectory is classified into The first unmatched logical trip set NPG1, and the second logical trips that do not match each of the first logical trips of the first trip trajectory are classified into the second unmatched logical trip set NPG2. Here, non-matching means that the above-mentioned conditions are not satisfied. In one example, if Equation 1 above is not satisfied, the i-th first logical run and the j-th second logical run do not match. In another example, if the "ith first logical trip's pit stop

and outbound sites

and outbound station”, then the i-th first logical trip and the j-th second logical trip do not match. In yet another example, if Equation 1 above is not satisfied, and the “i-th first logical trip” is not satisfied A pit stop for a logical trip

and outbound sites

and outbound station", the ith first logical trip and the jth second logical trip do not match.

在框S206中，获取匹配逻辑出行集中的各个第一逻辑出行的出行时间段和各自匹配的第二逻辑出行的出行时间段重叠的出行时间段，以得到各个重叠的出行时间段所对应的重叠时长。In block S206, obtain the travel time period overlapping the travel time period of each first logical trip in the matching logical travel set and the travel time period of each matching second logical trip, so as to obtain the overlap corresponding to each overlapping travel time period duration.

在一个示例中，以第i个第一逻辑出行和第j个第二逻辑出行相互匹配为例。例如，设定第i个第一逻辑出行包括早上8时到早上9时30分从进站站点S1经过站点S2、站点S3、站点S4而到达出站站点S5，而第j个第二逻辑出行包括早上8时到早上9时从进站站点S2经过站点S3而到达出站站点S4。那么第i个第一逻辑出行的出行时间段(早上8时到早上9时30分)和第j个第二逻辑出行的出行时间段(早上8时到早上9时)重叠的出行时间段为早上8时到早上9时，所对应的重叠时长为一个小时。In one example, the i-th first logical row and the j-th second logical row are matched with each other as an example. For example, it is assumed that the i-th first logical trip includes from 8:00 a.m. to 9:30 a.m. from inbound station S1 through station S2, station S3, and station S4 to outbound station S5, and the jth second logical trip Including from 8:00 am to 9:00 am from the inbound station S2 through the station S3 to the outbound station S4. Then the overlapping travel time period of the i-th first logical trip (8:00 a.m. to 9:30 a.m.) and the j-th second logical trip’s travel time period (8:00 a.m. to 9:00 a.m.) is From 8:00 am to 9:00 am, the corresponding overlapping time is one hour.

在这种情况下，可以得到匹配逻辑出行集PG中所有的第一逻辑出行各自对应的一个重叠时长。In this case, an overlapping duration corresponding to all the first logical trips in the matching logical trip set PG can be obtained.

在框S208中，对各个重叠时长进行衰减处理，以得到各个衰减处理后的重叠时长。In block S208, attenuation processing is performed on each overlapping duration to obtain each overlapping duration after attenuation processing.

对于“衰减处理”而言，可以包括对重叠时长进行衰减的衰减处理，也可以包括对重叠时长不进行衰减的衰减处理。The "attenuation processing" may include attenuation processing that attenuates the overlapping duration, or may include attenuation processing that does not attenuate the overlapping duration.

此外，在一个示例中，对匹配逻辑出行集PG中符合预设条件的各个第一逻辑出行所对应的重叠时长进行衰减，以得到符合预设条件的各个第一逻辑出行各自所对应的衰减重叠时长。言外之意，对匹配逻辑出行集PG中不符合预设条件的各个第一逻辑出行所对应的重叠时长不进行衰减。In addition, in an example, the overlapping duration corresponding to each first logical trip that meets the preset condition in the matching logical trip set PG is attenuated, so as to obtain the attenuation overlap corresponding to each first logical trip that meets the preset condition. duration. In other words, the overlapping duration corresponding to each first logical trip that does not meet the preset condition in the matching logical trip set PG is not attenuated.

当然，在一个示例中，预设条件的设置可以使匹配逻辑出行集PG中所有的第一逻辑出行所对应的重叠时长都被进行衰减。然而，在另一示例中，预设条件的设置可以使匹配逻辑出行集PG中部分的第一逻辑出行所对应的重叠时长被进行衰减，而其余部分的第一逻辑出行所对应的重叠时长不被进行衰减。Of course, in an example, the setting of the preset condition can cause the overlapping durations corresponding to all the first logical trips in the matching logical trip set PG to be attenuated. However, in another example, the setting of the preset condition can make the overlapping duration corresponding to some first logical trips in the matching logical trip set PG attenuated, while the overlapping duration corresponding to the remaining part of the first logical trips is not. is attenuated.

在一个示例中，所述预设条件可以包括：在匹配逻辑出行集PG中的一个出行时间段中，对应该出行时间段的所有第一出行轨迹的出行时长之和大于预设时长。以第一乘客出行30日为例，在第一乘客的第一逻辑出行中选取一个出行时间段，例如早上8时到早上9时的出行时间段，其从进站站点D1到达出站站点D2。在这个出行时间段，如果第一乘客30日中每日早上8时到早上9时，其都从进站站点D1到达出站站点D2，那么该出行时间段对应的所有第一逻辑出行(即30个第一逻辑出行)的出行时长为30小时；如果第一乘客在这30日中的20日的每日早上8时到早上9时，其都从进站站点D1到达出站站点D2，那么该出行时间段对应的所有第一逻辑出行(即20个第一逻辑出行)的出行时长为20小时。因此，设定一预设时长(诸如20小时等)，获取到对应匹配逻辑出行集PG中的某个出行时间段中的所有第一逻辑出行，如果这些第一逻辑出行的出行时长之和大于该预设时长，那么该出行时间段对应的这些第一逻辑出行符合所述预设条件，从而这些第一逻辑出行所对应的重叠时长都被进行衰减。In an example, the preset condition may include: in a travel time period in the matching logical travel set PG, the sum of the travel durations of all the first travel trajectories corresponding to the travel time period is greater than the preset time period. Taking the first passenger's trip for 30 days as an example, select a travel time period in the first logical trip of the first passenger, for example, the travel time period from 8:00 am to 9:00 am, which goes from the inbound site D1 to the outbound site D2 . In this travel time period, if the first passenger travels from the inbound station D1 to the outbound station D2 from 8:00 a.m. to 9:00 a.m. every day for 30 days, then all the first logical trips corresponding to this travel time period (that is, The travel duration of 30 first logical trips) is 30 hours; if the first passenger travels from the inbound station D1 to the outbound station D2 every day from 8:00 am to 9:00 am on the 20th of the 30 days, Then, the travel duration of all the first logical trips (ie, 20 first logical trips) corresponding to the travel time period is 20 hours. Therefore, a preset duration (such as 20 hours, etc.) is set to obtain all the first logical trips in a certain trip time period in the corresponding matching logical trip set PG, if the sum of the trip durations of these first logical trips is greater than For the preset duration, the first logical trips corresponding to the trip time period meet the preset condition, so that the overlapping durations corresponding to the first logical trips are all attenuated.

应当说明的是，匹配逻辑出行集PG中可以包括多个不同的出行时间段(或称出行模式)，这是由乘客的规律的出行时间段不同所导致的。It should be noted that the matching logical travel set PG may include multiple different travel time periods (or travel modes), which are caused by the different regular travel time periods of passengers.

此外，在上面的示例中，所述预设条件也可以按照天数来设定。例如，第一乘客在早上8时到早上9时，其都从进站站点D1到达出站站点D2，如果第一乘客在早上8时到早上9时的出行天数达到预设天数(例如20日)，那么认为早上8时到早上9时所对应的第一逻辑出行符合预设条件。In addition, in the above example, the preset condition may also be set according to the number of days. For example, if the first passenger travels from the inbound station D1 to the outbound station D2 from 8:00 am to 9:00 am, if the number of days the first passenger travels from 8:00 am to 9:00 am reaches a preset number of days (for example, 20 days) ), then it is considered that the first logical trip corresponding to 8:00 am to 9:00 am meets the preset condition.

在一个示例中，可以利用下面的式子2对匹配逻辑出行集PG中符合预设条件的各个第一逻辑出行所对应的重叠时长进行衰减。In an example, the following formula 2 can be used to attenuate the overlapping duration corresponding to each first logical trip that meets the preset condition in the matching logical trip set PG.

[2]

其中，a表示匹配逻辑出行集PG中的符合所述预设条件的第a个出行时间段，c _a表示所述第a个出行时间段中的第一逻辑出行的总数量，b表示第a个出行时间段中的第b个第一逻辑出行，γ表示衰减贡献率，

表示所述第b个第一逻辑出行所对应的重叠时长，

表示所述第b个第一逻辑出行所对应的衰减重叠时长。在这种情况下，通过对匹配逻辑出行集PG中符合所述预设条件的重叠时长进行衰减，降低符合所述预设条件的重叠时长在轨迹相似度计算过程中的权重，即降低了重叠出行时间段的权重(相对提高了非重叠出行时间段的比重)，从而提高不同数据源下同一乘客的识别匹配准确率。 Among them, a represents the a-th travel time period in the matching logical travel set PG that meets the preset condition, c _a represents the total number of first logical trips in the a-th travel time period, and b represents the a-th travel time period The b-th first logical trip in the travel time period, γ represents the decay contribution rate,

represents the overlapping duration corresponding to the bth first logical trip,

Indicates the decay overlap duration corresponding to the bth first logical row. In this case, by attenuating the overlapping durations that meet the preset conditions in the matching logical trip set PG, the weight of the overlapping durations that meet the preset conditions in the trajectory similarity calculation process is reduced, that is, the overlap is reduced. The weight of the travel time period (relatively increases the proportion of non-overlapping travel time periods), thereby improving the identification and matching accuracy of the same passenger under different data sources.

在一个示例中，将得到的各个第一逻辑出行各自所对应的衰减重叠时长以及匹配逻辑出行集PG中不符合所述预设条件的各个第一逻辑出行各自所对应的重叠时长确定为各个衰减处理后的重叠时长。也就是说，所有的衰减处理后的重叠时长包括各个对重叠时长进行衰减后的衰减重叠时长以及各个没有被衰减的重叠时长。In an example, the obtained attenuation overlap duration corresponding to each first logical trip and the overlapping duration corresponding to each first logical trip that does not meet the preset condition in the matching logical trip set PG are determined as the respective attenuations The duration of the overlap after processing. That is to say, all the overlapping durations after the attenuation process include the respective attenuated overlapping durations after the overlapping durations are attenuated and the respective overlapping durations that are not attenuated.

在框S210中，根据衰减处理后的重叠时长计算出第一出行轨迹和第二出行轨迹的轨迹相似度。In block S210, the track similarity between the first travel track and the second travel track is calculated according to the overlapping duration after the attenuation process.

在具体实现框S210的一个示例中，根据各个衰减处理后的重叠时长、第一出行轨迹的所有第一逻辑出行的出行最短时长以及与第一出行轨迹的所有第一逻辑出行都不匹配的第二逻辑出行(即第二未匹配逻辑出行集NPG2中的所有第二逻辑出行)的出行最短时长，计算出第一出行轨迹和第二出行轨迹的轨迹相似度。In an example of implementing block S210, according to the overlapping duration of each attenuation process, the shortest trip duration of all the first logical trips of the first trip trajectory, and the first trip trajectory that does not match all the first logical trips of the first trip trajectory The shortest travel duration of the second logical trip (that is, all the second logical trips in the second unmatched logical trip set NPG2), and the trajectory similarity between the first trip trajectory and the second trip trajectory is calculated.

在一个示例中，可以基于下面的式子3计算出第一出行轨迹和第二出行轨迹的轨迹相似度。In one example, the trajectory similarity of the first travel trajectory and the second travel trajectory may be calculated based on the following formula 3.

[3]

[3]

TS _afc表示第一出行轨迹，TS _ap表示第二出行轨迹，Sim(TS _afc,TS _ap)表示第一出行轨迹和第二出行轨迹的轨迹相似度，M表示匹配逻辑出行集PG中的符合预设条件的不同出行时间段的数量，α _p表示匹配逻辑出行集PG中的未符合预设条件的第p个所述第一逻辑出行对应的重叠时长，P表示匹配逻辑出行集PG中的不符合预设条件的第一逻辑出行的总数量，

表示第一出行轨迹中的第n个第一逻辑出行，

表示所述第n个第一逻辑出行的出行最短时长，N表示第一出行轨迹中的第一逻辑出行的总数量，

表示第二未匹配逻辑出行集NPG2中的第q个第二逻辑出行，

表示所述第q个第二逻辑出行的出行最短时长，Q表示第二未匹配逻辑出行集NPG2中的第二逻辑出行的总数量。 TS _afc indicates the first travel trajectory, TS _ap indicates the second travel trajectory, Sim(TS _afc ,TS _ap ) indicates the trajectory similarity between the first travel trajectory and the second travel trajectory, and M indicates the matching logic travel set PG in the pre-defined travel trajectory. Set the number of different travel time periods of the condition, α _p represents the overlapping duration corresponding to the p-th first logical trip in the matching logical trip set PG that does not meet the preset conditions, and P represents the non-matching logical trip set PG in the matching logical trip set PG. the total number of first logical trips that meet the preset conditions,

represents the nth first logical trip in the first trip trajectory,

represents the shortest trip duration of the nth first logical trip, N represents the total number of first logical trips in the first trip trajectory,

represents the qth second logical trip in the second unmatched logical trip set NPG2,

represents the shortest trip duration of the qth second logical trip, and Q represents the total number of second logical trips in the second unmatched logical trip set NPG2.

由上述可以知道，第一出行轨迹包括N个第一逻辑出行，而M和P均小于N，并且Q也小于第二出行轨迹中第二逻辑出行的总数量。It can be known from the above that the first travel trajectory includes N first logical trips, while M and P are both smaller than N, and Q is also smaller than the total number of second logical trips in the second travel trajectory.

在一个示例中，以第一乘客出行数据源为AFC数据源，第二乘客出行数据源为AP数据源为例，针对匹配逻辑出行集PG、第一未匹配逻辑出行集NPG1、第二未匹配逻辑出行集NPG2，可以分别对应三类出行情况。In an example, taking the first passenger travel data source as the AFC data source and the second passenger travel data source as the AP data source, for the matching logical travel set PG, the first unmatched logical travel set NPG1, the second unmatched logical travel set The logical travel set NPG2 can correspond to three types of travel situations.

具体地，匹配逻辑出行集PG所对应的第一类出行指的是乘客在进出站时被AFC设备所感测到，并且在出行过程中乘客所携带的移动终端被AP设备所检测到；第一未匹配逻辑出行集NPG1所对应的第二类出行指的是乘客在进出站时被AFC设备所感测到，但是在出行过程中乘客所携带的移动终端没有被AP设备所检测到；第二未匹配逻辑出行集NPG2所对应的第三类出行指的是乘客在进出站时未使用能够被AFC设备所感测到被感测对象(诸如公交卡等)，或者乘客只是在公共交通站附近活动，或者还可以是路过公共交通站却被AP设备所检测到。Specifically, the first type of trip corresponding to the matching logical trip set PG refers to that the passenger is detected by the AFC device when entering or leaving the station, and the mobile terminal carried by the passenger is detected by the AP device during the trip; the first The second type of travel corresponding to the unmatched logical travel set NPG1 refers to that the passenger is detected by the AFC device when entering or leaving the station, but the mobile terminal carried by the passenger is not detected by the AP device during the travel process; The third type of travel corresponding to the matching logical travel set NPG2 refers to that the passenger does not use the sensed object (such as a bus card, etc.) that can be sensed by the AFC device when entering or leaving the station, or the passenger is only active near the public transportation station, Or it can also be detected by the AP device when passing a public transportation station.

在实际过程中，考虑这三类出行情况各不相同，因此在计算轨迹相似度时分别给予不同的惩罚系数(也可以是权重系数，即如果惩罚系数大，相应的权重系数小)。例如，第一类出行和第二类出行发生的可能性大，所以相应的惩罚系数较小(即权重系数较大)，而第三类出行发生的可能性小，所以相应的惩罚系数较大(即权重系数较小)。通过这样的设计，可以有效降低不相关的出行轨迹的干扰。In the actual process, considering that these three types of travel situations are different, different penalty coefficients are given respectively when calculating the similarity of the trajectory (it can also be a weight coefficient, that is, if the penalty coefficient is large, the corresponding weight coefficient is small). For example, the first type of trip and the second type of trip are more likely to occur, so the corresponding penalty coefficient is small (that is, the weight coefficient is large), while the third type of trip is less likely to occur, so the corresponding penalty coefficient is large. (that is, the weight coefficient is smaller). Through such a design, the interference of irrelevant travel trajectories can be effectively reduced.

因此，基于上述分析，在具体实现框S210的另一个示例中，分别向匹配逻辑出行集PG中的第一逻辑出行的出行最短时长、第一未匹配逻辑出行集 NPG1中的第一逻辑出行的出行最短时长、第二未匹配逻辑出行集NPG2中的第二逻辑出行的出行最短时长赋予不同的惩罚系数。Therefore, based on the above analysis, in another example of implementing block S210, the shortest travel duration of the first logical trip in the matching logical trip set PG and the trip duration of the first logical trip in the first unmatched logical trip set NPG1 are respectively directed to The shortest trip duration and the shortest trip duration of the second logical trip in the second unmatched logical trip set NPG2 are given different penalty coefficients.

进一步地，根据各个衰减处理后的重叠时长，并根据被分别赋予了不同的惩罚系数的匹配逻辑出行集PG中的第一逻辑出行的出行最短时长、第一未匹配逻辑出行集NPG1中的第一逻辑出行的出行最短时长、第二未匹配逻辑出行集NPG2中的第二逻辑出行的出行最短时长，计算出第一出行轨迹和第二出行轨迹的相似度。具体地，可以基于下面的式子4计算出第一出行轨迹和第二出行轨迹的相似度。Further, according to the overlapping durations after each attenuation process, and according to the shortest travel duration of the first logical trip in the matching logical trip set PG, which are respectively given different penalty coefficients, and the first unmatched logical trip set NPG1. The shortest trip duration of a logical trip, the shortest trip duration of the second logical trip in the second unmatched logical trip set NPG2, and the similarity between the first trip trajectory and the second trip trajectory is calculated. Specifically, the similarity between the first travel trajectory and the second travel trajectory can be calculated based on the following formula 4.

[4]

[4]

其中，TS _afc表示第一出行轨迹，TS _ap表示第二出行轨迹，Sim(TS _afc,TS _ap)表示第一出行轨迹和第二出行轨迹的轨迹相似度，M表示匹配逻辑出行集PG中的符合预设条件的不同出行时间段的总数量，α _p表示匹配逻辑出行集PG中的未符合预设条件的第p个第一逻辑出行对应的重叠时长，P表示匹配逻辑出行集PG中的未符合预设条件的第一逻辑出行的总数量，

表示匹配逻辑出行集PG中的第s个第一逻辑出行，

表示所述第s个第一逻辑出行的出行最短时长，S表示匹配逻辑出行集PG中的第一逻辑出行的总数量，

表示第一未匹配逻辑出行集NPG1中的第r个第一逻辑出行，

表示所述第r个第一逻辑出行的出行最短时长，R表示第一未匹配逻辑出行集NPG1中的第一逻辑出行的总数量，

表示第二未匹配逻辑出行集NPG2中的第q个第二逻辑出行，

表示所述第q个第二逻辑出行的出行最短时长，Q表示第二未匹配逻辑出行集NPG2中的第二逻辑出行的总数量，θ ₁、θ ₂和θ ₃分别表示不同的惩罚系数。 Among them, TS _afc represents the first travel trajectory, TS _ap represents the second travel trajectory, Sim(TS _afc , TS _ap ) represents the trajectory similarity between the first travel trajectory and the second travel trajectory, and M represents the matching logical travel set PG. The total number of different travel time periods that meet the preset conditions, α _p represents the overlapping duration corresponding to the p-th first logical trip that does not meet the preset conditions in the matching logical trip set PG, and P represents the matching logical trip set PG. the total number of first logical trips that do not meet the preset conditions,

represents the sth first logical trip in the matching logical trip set PG,

represents the shortest travel duration of the s-th first logical trip, S represents the total number of first logical trips in the matching logical trip set PG,

represents the rth first logical trip in the first unmatched logical trip set NPG1,

represents the shortest trip duration of the rth first logical trip, R represents the total number of first logical trips in the first unmatched logical trip set NPG1,

represents the shortest trip duration of the qth second logical trip, Q represents the total number of second logical trips in the second unmatched logical trip set NPG2, and θ ₁ , θ ₂ and θ ₃ respectively represent different penalty coefficients.

由上述可以知道，第一出行轨迹包括S+R(二者之和等于式子3中的N)个第一逻辑出行，而M和P均小于S+R，并且Q也小于第二出行轨迹中第二逻辑出行的总数量。It can be known from the above that the first travel trajectory includes S+R (the sum of the two is equal to N in Equation 3) first logical travels, while M and P are both smaller than S+R, and Q is also smaller than the second travel trajectory. The total number of second logical trips in .

在框S212中，根据轨迹相似度来对第一乘客和第二乘客进行识别。当然，这里的识别指的是识别第一乘客和第二乘客是否是同一乘客。In block S212, the first passenger and the second passenger are identified according to the trajectory similarity. Of course, the identification here refers to identifying whether the first passenger and the second passenger are the same passenger.

在一个示例中，可以设置一轨迹相似度阈值，而该轨迹相似度阈值可以是指定的。在这种情况下，当计算出的轨迹相似度大于或者等于该轨迹相似度阈值时，识别出第一乘客和第二乘客是同一乘客；而当计算出的轨迹相似度小于该轨迹相似度阈值时，识别出第一乘客和第二乘客不是同一乘客。In one example, a trajectory similarity threshold may be set, and the trajectory similarity threshold may be specified. In this case, when the calculated trajectory similarity is greater than or equal to the trajectory similarity threshold, the first passenger and the second passenger are identified as the same passenger; and when the calculated trajectory similarity is less than the trajectory similarity threshold , it is recognized that the first passenger and the second passenger are not the same passenger.

图3A是根据本发明的实施例的公共交通乘客的识别装置的流程图。FIG. 3A is a flowchart of an apparatus for identifying a public transportation passenger according to an embodiment of the present invention.

参照图3A，根据本发明的实施例的公共交通乘客的识别装置300包括：出行轨迹获取模块310、逻辑出行匹配模块320、重叠时长获取模块330、重叠时长衰减模块340、轨迹相似度计算模块350以及乘客识别模块360。Referring to FIG. 3A , an apparatus 300 for identifying public transportation passengers according to an embodiment of the present invention includes: a travel trajectory acquisition module 310 , a logical travel matching module 320 , an overlapping duration acquisition module 330 , an overlapping duration attenuation module 340 , and a trajectory similarity calculation module 350 And the passenger identification module 360 .

出行轨迹获取模块310被配置为基于第一乘客出行数据源获取包括多个第一逻辑出行的第一出行轨迹，以及基于第二乘客出行数据源获取包括多个第二逻辑出行的第二出行轨迹。在一个示例中，每个第一逻辑出行包括对应的一个进站站点、一个进站时间、一个出站站点、一个出站时间、一个出行时间段、一个出行时长以及一个最短出行时长。每个第二逻辑出行包括对应的一个进站站点(按照时间先后顺序排列的第一个途径站点)、一个进站时间(第一个途径站点对应的途径时间)、一个出站站点(按照时间先后顺序排列的最后一个途径站点)、一个出站时间(最后一个途径站点对应的途径时间)、一个出行时间段、一个出行时长以及一个最短出行时长。The travel trajectory obtaining module 310 is configured to obtain a first travel trajectory including a plurality of first logical trips based on the first passenger travel data source, and obtain a second travel trajectory including a plurality of second logical trips based on the second passenger travel data source . In one example, each first logical trip includes a corresponding inbound station, an inbound time, an outbound station, an outbound time, a travel time period, a travel duration, and a minimum travel duration. Each second logical trip includes a corresponding inbound station (the first route station in chronological order), an inbound time (the route time corresponding to the first route station), and an outbound station (by time The last route station in the sequence), a departure time (the route time corresponding to the last route station), a travel time period, a travel time, and a shortest travel time.

逻辑出行匹配模块320被配置为对各个第一逻辑出行和各个第二逻辑出行进行匹配，以将与第二逻辑出行匹配的第一逻辑出行归类于匹配逻辑出行集。The logical trip matching module 320 is configured to match each of the first logical trips and each of the second logical trips to classify the first logical trips matching the second logical trips into a set of matching logical trips.

在一个示例中，参照图3B，逻辑出行匹配模块320可以包括：进出站时长确定单元321、最短进出站时长确定单元322以及匹配确定单元323。所述进出站时长确定单元321被配置为利用第i个第一逻辑出行的进站时间

减去第j个第二逻辑出行的进站时间

以得到进站时长，以及利用第j个第二逻辑出行的出站时间

减去第i个第一逻辑出行的出站时间

以得到出站时长。所述最短进出站时长确定单元322被配置为获取从第i个第一逻辑出行的进站站点

到第j个第二逻辑出行的进站站点

之间的最短进站时长

且获取从第j个第二逻辑出行的出站站点

到第i个第一逻辑出行的出站站点

之间的最短出站时长

所述匹配确定单元323被配置为如果所述进站时长和所述最短进站时长

之和的绝对值小于所述预设时长阈值ε'，则认为第i个第一逻辑出行和第j个第二逻辑出行相互匹配，并将与第j个第二逻辑出行匹配的第i个第一逻辑出行归类于匹配逻辑出行集PG。因此，在这个示例中，所述匹配确定单元323可以根据上面的式子1来确定第i个第一逻辑出行和第j个第二逻辑出行相互匹配。 In one example, referring to FIG. 3B , the logical travel matching module 320 may include: an inbound and outbound duration determination unit 321 , a shortest inbound and outbound duration determination unit 322 , and a matching determination unit 323 . The inbound and outbound duration determination unit 321 is configured to use the inbound time of the i-th first logical trip

Subtract the pit stop time of the jth second logical trip

Subtract the outbound time of the ith first logical trip

to get the outbound time. The shortest inbound and outbound duration determination unit 322 is configured to obtain the inbound site from the i-th first logical trip

to the pit stop of the jth second logical trip

Minimum pit stop time between

and get the outbound stop from the jth second logical trip

To the outbound stop of the i-th first logical trip

Minimum outbound time between

The matching determination unit 323 is configured to determine if the pit stop duration and the shortest pit stop duration

The absolute value of the sum is less than the preset duration threshold ε', then it is considered that the i-th first logical trip and the j-th second logical trip match each other, and the i-th logical trip that matches the j-th second logical trip The first logical trip is classified into the matching logical trip set PG. Therefore, in this example, the matching determination unit 323 may determine that the i-th first logical row and the j-th second logical row match each other according to the above formula 1.

此外，在另一个示例中，所述匹配确定单元323还可以根据条件“第i个第一逻辑出行的进站站点

和出站站点

之间的有效路径经过第j个第二逻辑出行的进站站点

和出站站点

”来确定第i个第一逻辑出行和第j个第二逻辑出行相互匹配。 In addition, in another example, the matching determination unit 323 may also be based on the condition "the pit stop of the i-th first logical trip

and outbound sites

” to determine that the i-th first logical line and the j-th second logical line match each other.

另外，在一个示例中，对各个第一逻辑出行和各个第二逻辑出行进行匹配之后，将与第二出行轨迹的各个第二逻辑出行都不匹配的第一逻辑出行归类于第一未匹配逻辑出行集NPG1，并将与第一出行轨迹的各个第一逻辑出行都不匹配的第二逻辑出行归类于第二未匹配逻辑出行集NPG2。这里，不匹配指的是不满足上述的条件。在一个示例中，如果不满足上面的式子1，则第i个第一逻辑出行和第j个第二逻辑出行不匹配。在另一示例中，如果不满足“第i个第一逻辑出行的进站站点

和出站站点

之间的有效路径经过第j个第二逻辑出行的进站站点

和出站站点

之间的有效路径经过第j个第二逻辑出行的进站站点

和出站站点”，则第i个第一逻辑出行和第j个第二逻辑出行不匹配。 In addition, in one example, after each first logical trip and each second logical trip are matched, the first logical trip that does not match each second logical trip of the second trip trajectory is classified as the first unmatched The logical travel set NPG1 is set, and the second logical travel that does not match each of the first logical travels of the first travel trajectory is classified into the second unmatched logical travel set NPG2. Here, mismatch means that the above-mentioned conditions are not satisfied. In one example, if Equation 1 above is not satisfied, the i-th first logical run and the j-th second logical run do not match. In another example, if the "ith first logical trip's pit stop

and outbound sites

重叠时长获取模块330被配置为获取匹配逻辑出行集中的各个第一逻辑出行的出行时间段和各自匹配的第二逻辑出行的出行时间段重叠的出行时间段，以得到各个重叠的出行时间段所对应的重叠时长。The overlapping duration obtaining module 330 is configured to obtain the travel time period overlapping the travel time period of each first logical trip in the matching logical travel set and the travel time period of the respective matching second logical trip, so as to obtain the total length of each overlapping travel time period. The corresponding overlap duration.

重叠时长衰减模块340被配置为对各个重叠时长进行衰减处理，以得到各个衰减处理后的重叠时长。The overlapping duration attenuation module 340 is configured to perform attenuation processing on each overlapping duration to obtain each attenuated overlapping duration.

针对“衰减处理”而言，其可以包括对重叠时长进行衰减的衰减处理，也可以包括对重叠时长不进行衰减的衰减处理。For "attenuation processing", it may include attenuation processing that attenuates the overlapping duration, or may include attenuation processing that does not attenuate the overlapping duration.

在一个示例中，参照图3C，重叠时长衰减模块340包括时长衰减单元341和时长确定单元342。所述时长衰减单元341被配置为对匹配逻辑出行集PG中符合预设条件的各个第一逻辑出行所对应的重叠时长进行衰减，以得到符合预设条件的各个第一逻辑出行各自所对应的衰减重叠时长。言外之意，对匹配逻辑出行集PG中不符合预设条件的各个第一逻辑出行所对应的重叠时长不进行衰减。时长确定单元342被配置为将得到的各个第一逻辑出行各自所对应的衰减重叠时长以及匹配逻辑出行集PG中不符合所述预设条件的各个第一逻辑出行各自所对应的重叠时长确定为各个衰减处理后的重叠时长。也就是说，所有的衰减处理后的重叠时长包括各个对重叠时长进行衰减后的衰减重叠时长以及各个没有被衰减的重叠时长。In one example, referring to FIG. 3C , the overlapping duration attenuation module 340 includes a duration attenuation unit 341 and a duration determination unit 342 . The duration attenuation unit 341 is configured to attenuate the overlapping durations corresponding to the first logical trips that meet the preset conditions in the matching logical trip set PG, so as to obtain the corresponding values of the first logical trips that meet the preset conditions. Decay overlap duration. In other words, the overlapping duration corresponding to each first logical trip that does not meet the preset condition in the matching logical trip set PG is not attenuated. The duration determining unit 342 is configured to determine the obtained attenuation overlap duration corresponding to each first logical trip and the overlapping duration corresponding to each first logical trip that does not meet the preset condition in the matching logical trip set PG as: The duration of the overlap after each decay process. That is to say, all the overlapping durations after the attenuation process include the respective attenuated overlapping durations after the overlapping durations are attenuated and the respective overlapping durations that are not attenuated.

在一个示例中，所述时长衰减单元341可以利用上面的式子2对匹配逻辑出行集PG中符合预设条件的各个第一逻辑出行所对应的重叠时长进行衰减。In an example, the duration attenuation unit 341 may use the above formula 2 to attenuate the overlapping durations corresponding to each first logical trip that meets the preset condition in the matching logical trip set PG.

轨迹相似度计算模块350被配置为根据衰减处理后的重叠时长计算出第一出行轨迹和第二出行轨迹的轨迹相似度。The trajectory similarity calculation module 350 is configured to calculate the trajectory similarity of the first travel trajectory and the second travel trajectory according to the overlapping duration after attenuation processing.

在一个示例中，轨迹相似度计算模块350进一步被配置为根据各个衰减处理后的重叠时长、第一出行轨迹的所有第一逻辑出行的出行最短时长以及与第一出行轨迹的所有第一逻辑出行都不匹配的第二逻辑出行(即第二未匹配逻辑出行集NPG2中的所有第二逻辑出行)的出行最短时长，计算出第一出行轨迹和第二出行轨迹的轨迹相似度。在这种情况下，轨迹相似度计算模块350可以基于上面的式子3计算出第一出行轨迹和第二出行轨迹的轨迹相似度。In one example, the trajectory similarity calculation module 350 is further configured to, according to the respective attenuated overlapping durations, the shortest trip durations of all first logical trips of the first trip trajectory, and all first logical trips related to the first trip trajectory According to the shortest travel duration of the second logical trips that do not match (ie, all the second logical trips in the second unmatched logical trip set NPG2), the trajectory similarity between the first trip trajectory and the second trip trajectory is calculated. In this case, the trajectory similarity calculation module 350 may calculate the trajectory similarity of the first travel trajectory and the second travel trajectory based on the above formula 3.

在另一个示例中，参照图3D，轨迹相似度计算模块350可以包括权重系数赋予单元351、相似度计算单元352。所述权重系数赋予单元351被配置为分别向匹配逻辑出行集PG中的第一逻辑出行的出行最短时长、第一未匹配逻辑出行集NPG1中的第一逻辑出行的出行最短时长、第二未匹配逻辑出行集 NPG2中的第二逻辑出行的出行最短时长赋予不同的惩罚系数。所述相似度计算单元352被根据各个衰减处理后的重叠时长，并根据被分别赋予了不同的惩罚系数的匹配逻辑出行集PG中的第一逻辑出行的出行最短时长、第一未匹配逻辑出行集NPG1中的第一逻辑出行的出行最短时长、第二未匹配逻辑出行集NPG2中的第二逻辑出行的出行最短时长，计算出第一出行轨迹和第二出行轨迹的相似度。在这种情况下，该相似度计算单元352可以基于上面的式子4计算出第一出行轨迹和第二出行轨迹的相似度。In another example, referring to FIG. 3D , the trajectory similarity calculation module 350 may include a weight coefficient assignment unit 351 and a similarity calculation unit 352 . The weight coefficient assigning unit 351 is configured to assign the shortest travel duration of the first logical trip in the matching logical trip set PG, the shortest trip duration of the first logical trip in the first unmatched logical trip set NPG1, and the shortest trip duration of the second logical trip in the first unmatched logical trip set NPG1. The shortest travel duration of the second logical trip in the matching logical trip set NPG2 is given different penalty coefficients. The similarity calculation unit 352 is based on the overlapping durations after each attenuation process, and according to the shortest trip duration of the first logical trip and the first unmatched logical trip in the matching logical trip set PG that have been assigned different penalty coefficients respectively. The shortest trip duration of the first logical trip in the set NPG1 and the shortest trip duration of the second logical trip in the second unmatched logical trip set NPG2 are used to calculate the similarity between the first trip trajectory and the second trip trajectory. In this case, the similarity calculation unit 352 may calculate the similarity between the first travel trajectory and the second travel trajectory based on the above formula 4.

乘客识别模块360被配置为根据轨迹相似度来对第一乘客和第二乘客进行识别。当然，这里的识别指的是识别第一乘客和第二乘客是否是同一乘客。The passenger identification module 360 is configured to identify the first passenger and the second passenger according to the trajectory similarity. Of course, the identification here refers to identifying whether the first passenger and the second passenger are the same passenger.

参照图4，电子设备400可以包括至少一个处理器410、存储器(例如，非易失性存储器)420、内存430和通信接口440，并且至少一个处理器410、存储器420、内存430和通信接口440经由总线450连接在一起。至少一个处理器410执行在存储器中存储或编码的至少一个机器可读指令(即，上述以软件形式实现的元素)。4 , an electronic device 400 may include at least one processor 410 , memory (eg, non-volatile memory) 420 , memory 430 , and communication interface 440 , and at least one processor 410 , memory 420 , memory 430 , and communication interface 440 Connected together via bus 450 . At least one processor 410 executes at least one machine-readable instruction stored or encoded in memory (ie, the above-described elements implemented in software).

在一个示例中，在存储器中存储计算机可执行指令，其当执行时使得至少一个处理器410执行以下过程：基于第一乘客出行数据源获取包括多个第一逻辑出行的第一出行轨迹，以及基于第二乘客出行数据源获取包括多个第二逻辑出行的第二出行轨迹，第一逻辑出行和第二逻辑出行分别包括构成一次逻辑出行的出行时间段；对第一逻辑出行和第二逻辑出行进行匹配，以将与第二逻辑出行匹配的第一逻辑出行归类于匹配逻辑出行集；获取匹配逻辑出行集中的第一逻辑出行的出行时间段和其匹配的第二逻辑出行的出行时间段重叠的出行时间段，以得到重叠的出行时间段所对应的重叠时长；对重叠时长进行衰减处理，以得到衰减处理后的重叠时长；根据衰减处理后的重叠时长计算出第一出行轨迹和第二出行轨迹的轨迹相似度；根据轨迹相似度来对第一乘客和第二乘客进行识别。当然，这里的识别指的是识别第一乘客和第二乘客是否是同一乘客。In one example, computer-executable instructions are stored in memory that, when executed, cause at least one processor 410 to perform a process of: obtaining a first travel trajectory including a plurality of first logical trips based on the first passenger travel data source, and A second travel trajectory including multiple second logical trips is obtained based on the second passenger travel data source, and the first logical trip and the second logical trip respectively include travel time periods that constitute one logical trip; Matching trips to classify the first logical trip that matches the second logical trip into the matching logical trip set; obtain the trip time period of the first logical trip in the matching logical trip set and the trip time of the matching second logical trip The overlapping travel time segments are used to obtain the overlapping duration corresponding to the overlapping travel time segments; the overlapping duration is attenuated to obtain the attenuated overlapping duration; the first travel trajectory and The track similarity of the second travel track; the first passenger and the second passenger are identified according to the track similarity. Of course, the identification here refers to identifying whether the first passenger and the second passenger are the same passenger.

应该理解，在存储器中存储的计算机可执行指令当执行时使得至少一个处理器410在进行根据本发明的实施例中结合以上图2描述的各种操作和功能。It should be understood that the computer-executable instructions stored in the memory, when executed, cause at least one processor 410 to perform the various operations and functions described above in connection with FIG. 2 in embodiments in accordance with the present invention.

根据一个实施例，提供了一种例如机器可读介质的程序产品。机器可读介质可以具有指令(即，上述以软件形式实现的元素)，该指令当被机器执行时，使得机器执行本发明的实施例中的结合以上图2描述的各种操作和功能。According to one embodiment, a program product, eg, a machine-readable medium, is provided. A machine-readable medium may have instructions (ie, the above-described elements implemented in software) that, when executed by a machine, cause the machine to perform the various operations and functions described in connection with FIG. 2 above in embodiments of the present invention.

具体地，可以提供配有可读存储介质的***或者装置，在该可读存储介质上存储着实现上述实施例中任一实施例的功能的软件程序代码，且使该***或者装置的计算机或处理器读出并执行存储在该可读存储介质中的指令。Specifically, a system or an apparatus equipped with a readable storage medium may be provided, on which software program codes for realizing the functions of any of the above-described embodiments are stored, and a computer or a computer of the system or apparatus may be provided. The processor reads and executes the instructions stored in the readable storage medium.

在这种情况下，从可读介质读取的程序代码本身可实现上述实施例中任何一项实施例的功能，因此机器可读代码和存储机器可读代码的可读存储介质构成了本发明的实施例的一部分。In this case, the program code itself read from the readable medium can implement the functions of any one of the above-described embodiments, and thus the machine-readable code and the readable storage medium storing the machine-readable code constitute the present invention part of the example.

可读存储介质的实施例包括软盘、硬盘、磁光盘、光盘(如CD-ROM、CD-R、CD-RW、DVD-ROM、DVD-RAM、DVD-RW、DVD-RW)、磁带、非易失性存储卡和ROM。可选择地，可以由通信网络从服务器计算机上或云上下载程序代码。Examples of readable storage media include floppy disks, hard disks, magneto-optical disks, optical disks (eg, CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD-RW), magnetic tape, non- Volatile memory cards and ROMs. Alternatively, the program code may be downloaded from a server computer or the cloud over a communications network.

上述对本发明的特定实施例进行了描述。其它实施例在所附权利要求书的范围内。在一些情况下，在权利要求书中记载的动作或步骤可以按照不同于实施例中的顺序来执行并且仍然可以实现期望的结果。另外，在附图中描绘的过程不一定要求示出的特定顺序或者连续顺序才能实现期望的结果。在某些实施方式中，多任务处理和并行处理也是可以的或者可能是有利的。The foregoing describes specific embodiments of the invention. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different from that in the embodiments and still achieve desirable results. Additionally, the processes depicted in the figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

上述各流程和各***结构图中不是所有的步骤和单元都是必须的，可以根据实际的需要忽略某些步骤或单元。各步骤的执行顺序不是固定的，可以根据需要进行确定。上述各实施例中描述的装置结构可以是物理结构，也可以是逻辑结构，即，有些单元可能由同一物理实体实现，或者，有些单元可能分由多个物理实体实现，或者，可以由多个独立设备中的某些部件共同实现。Not all steps and units in the above processes and system structure diagrams are necessary, and some steps or units may be omitted according to actual needs. The execution order of each step is not fixed and can be determined as required. The device structure described in the above embodiments may be a physical structure or a logical structure, that is, some units may be implemented by the same physical entity, or some units may be implemented by multiple physical entities, or may be implemented by multiple physical entities. Some components in separate devices are implemented together.

在整个本说明书中使用的术语“示例性”、“示例”等意味着“用作示例、实例或例示”，并不意味着比其它实施例“优选”或“具有优势”。出于提供对所描述技术的理解的目的，具体实施方式包括具体细节。然而，可以在没有这些具体细节的情况下实施这些技术。在一些实例中，为了避免对所描述的实施例的概念造成难以理解，公知的结构和装置以框图形式示出。The terms "exemplary", "example" and the like used throughout this specification mean "serving as an example, instance or illustration" and do not mean "preferred" or "advantage" over other embodiments. The detailed description includes specific details for the purpose of providing an understanding of the described technology. However, these techniques may be practiced without these specific details. In some instances, well-known structures and devices are shown in block diagram form in order to avoid obscuring the concepts of the described embodiments.

以上结合附图详细描述了本发明的实施例的可选实施方式，但是，本发明的实施例并不限于上述实施方式中的具体细节，在本发明的实施例的技术构思范围内，可以对本发明的实施例的技术方案进行多种简单变型，这些简单变型均属于本发明的实施例的保护范围。The optional embodiments of the embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details of the above-mentioned embodiments. Within the scope of the technical concept of the embodiments of the present invention, the The technical solutions of the embodiments of the present invention undergo various simple modifications, and these simple modifications all belong to the protection scope of the embodiments of the present invention.

本说明书内容的上述描述被提供来使得本领域任何普通技术人员能够实现或者使用本说明书内容。对于本领域普通技术人员来说，对本说明书内容进行的各种修改是显而易见的，并且，也可以在不脱离本说明书内容的保护范围的情况下，将本文所定义的一般性原理应用于其它变型。因此，本说明书内容并不限于本文所描述的示例和设计，而是与符合本文公开的原理和新颖性特征的最广范围相一致。The above description of the present specification is provided to enable any person of ordinary skill in the art to make or use the present specification. Various modifications to this specification will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other variations without departing from the scope of this specification . Thus, this disclosure is not intended to be limited to the examples and designs described herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

一种公共交通乘客的识别方法，其中，所述识别方法包括：A method for identifying public transport passengers, wherein the identifying method comprises:

基于第一乘客出行数据源获取包括多个第一逻辑出行的第一出行轨迹，以及基于第二乘客出行数据源获取包括多个第二逻辑出行的第二出行轨迹，所述第一逻辑出行和所述第二逻辑出行分别包括构成一次逻辑出行的出行时间段；A first travel trajectory including a plurality of first logical trips is obtained based on a first passenger travel data source, and a second travel trajectory including a plurality of second logical trips is obtained based on a second passenger travel data source, the first logical travel and The second logical trips respectively include travel time periods constituting one logical trip;

对所述第一逻辑出行和所述第二逻辑出行进行匹配，以将与所述第二逻辑出行匹配的所述第一逻辑出行归类于匹配逻辑出行集；matching the first logical trip and the second logical trip to classify the first logical trip matching the second logical trip into a set of matching logical trips;

获取所述匹配逻辑出行集中的所述第一逻辑出行的出行时间段和其匹配的所述第二逻辑出行的出行时间段重叠的出行时间段，以得到所述重叠的出行时间段所对应的重叠时长；Obtain the travel time period in which the travel time period of the first logical trip and the travel time period of the matching second logical trip in the matching logical travel set overlap, so as to obtain the corresponding travel time period of the overlapping travel time period. overlapping duration;

对所述重叠时长进行衰减处理，以得到衰减处理后的重叠时长；Perform attenuation processing on the overlapping duration to obtain the overlapping duration after attenuation processing;

根据所述衰减处理后的重叠时长计算出所述第一出行轨迹和所述第二出行轨迹的轨迹相似度；Calculate the trajectory similarity of the first travel trajectory and the second travel trajectory according to the overlapping duration after the attenuation process;

根据所述轨迹相似度来来对第一乘客和第二乘客进行识别。The first passenger and the second passenger are identified according to the trajectory similarity.
根据权利要求1所述的识别方法，其中，所述对所述重叠时长进行衰减处理，以得到衰减处理后的重叠时长，具体包括：The identification method according to claim 1, wherein the performing attenuation processing on the overlapping durations to obtain the overlapping durations after attenuation processing, specifically comprising:

对所述匹配逻辑出行集中符合预设条件的所述第一逻辑出行所对应的所述重叠时长进行衰减，以得到衰减重叠时长；attenuating the overlapping duration corresponding to the first logical trip that meets the preset condition in the matching logical trip set to obtain the attenuation overlapping duration;

将所述衰减重叠时长和所述匹配逻辑出行集中不符合所述预设条件的所述第一逻辑出行所对应的所述重叠时长确定为所述衰减处理后的重叠时长；Determining the attenuation overlap duration and the overlap duration corresponding to the first logical trip that does not meet the preset condition in the matching logical trip set as the attenuation-processed overlap duration;

其中，所述预设条件包括：在所述匹配逻辑出行集中的一个出行时间段中，对应该出行时间段的所有所述第一逻辑出行的出行时长之和大于预设时长。Wherein, the preset condition includes: in a travel time period in the matching logical travel set, the sum of travel time lengths of all the first logical travels corresponding to the travel time period is greater than a preset time period.
根据权利要求2所述的识别方法，其中，所述对所述匹配逻辑出行集中符合预设条件的所述第一逻辑出行所对应的所述重叠时长进行衰减，以得到衰减重叠时长，具体包括：基于下面的式子1计算得到所述衰减重叠时长，The identification method according to claim 2, wherein the attenuating the overlapping duration corresponding to the first logical trip that meets the preset condition in the matching logical trip set to obtain the attenuation overlapping duration, specifically comprises: : Calculate the decay overlap duration based on the following formula 1,

其中，a表示所述匹配逻辑出行集中的符合所述预设条件的第a个出行时间段，c _a表示所述第a个出行时间段中的所述第一逻辑出行的总数量，b表示第a个出行时间段中的第b个第一逻辑出行，γ表示衰减贡献率，
表示所述第b个第一逻辑出行所对应的重叠时长，
表示所述第b个第一逻辑出行所对应的衰减重叠时长。 Among them, a represents the a-th travel time period that meets the preset condition in the matching logical travel set, c _a represents the total number of the first logical trips in the a-th travel time period, and b represents The bth first logical trip in the ath trip time period, γ represents the decay contribution rate,
represents the overlapping duration corresponding to the bth first logical trip,
Indicates the decay overlap duration corresponding to the bth first logical row.
根据权利要求1至3任一项所述的识别方法，其中，所述第一逻辑出行和所述第二逻辑出行均还包括构成一次逻辑出行的出行最短时长；The identification method according to any one of claims 1 to 3, wherein both the first logical trip and the second logical trip further include a shortest trip duration that constitutes a logical trip;

其中，所述根据所述衰减处理后的重叠时长计算出所述第一出行轨迹和所述第二出行轨迹的轨迹相似度，具体包括：Wherein, calculating the trajectory similarity between the first travel trajectory and the second travel trajectory according to the overlapping duration after the attenuation process specifically includes:

根据所述衰减处理后的重叠时长、所述多个第一逻辑出行的出行最短时长以及与所述多个第一逻辑出行都不匹配的所述第二逻辑出行的出行最短时长，计算出所述第一出行轨迹和所述第二出行轨迹的轨迹相似度。According to the overlapping duration after the attenuation process, the shortest travel duration of the multiple first logical trips, and the shortest trip duration of the second logical trip that does not match the multiple first logical trips, calculate the track similarity between the first travel track and the second travel track.
根据权利要求4所述的识别方法，其中，所述根据所述衰减处理后的重叠时长、所述多个第一逻辑出行的出行最短时长以及与所述多个第一逻辑出行都不匹配的所述第二逻辑出行的出行最短时长，计算出所述第一出行轨迹和所述第二出行轨迹的轨迹相似度，具体包括：基于下面的式子2计算得到所述轨迹相似度，The identification method according to claim 4, wherein the overlapping duration processed according to the attenuation, the shortest duration of the trip of the plurality of first logical trips, and the trip duration that does not match the plurality of first logical trips The shortest travel duration of the second logical trip, and calculating the trajectory similarity of the first travel trajectory and the second travel trajectory, specifically including: calculating and obtaining the trajectory similarity based on the following formula 2,

其中，TS _afc表示所述第一出行轨迹，TS _ap表示所述第二出行轨迹，Sim(TS _afc,TS _ap)表示所述第一出行轨迹和所述第二出行轨迹的轨迹相似度，M表示所述匹配逻辑出行集中的符合预设条件的不同出行时间段的数量，α _p表示所述匹配逻辑出行集中的未符合预设条件的第p个所述第一逻辑出行对应的重叠时长，P表示所述匹配逻辑出行集中的不符合预设条件的第一逻辑出行的总数量，
表示所述第一出行轨迹中的第n个第一逻辑出行，
表示所述第n个第一逻辑出行的出行最短时长，N表示所述第一出行轨迹中的所述第一逻辑出行的总数量，
表示与所述第一出行轨迹的各个第一逻辑出行都不匹配的第二逻辑出行中的第q个第二逻辑出行，
表示所述第q个第二逻辑出行的出行最短时长，Q表示与所述第一出行轨迹的各个第一逻辑出行都不匹配的第二逻辑出行的总数量。 Wherein, TS _afc represents the first travel trajectory, TS _ap represents the second travel trajectory, Sim(TS _afc , TS _ap ) represents the trajectory similarity between the first travel trajectory and the second travel trajectory, M represents the number of different travel time periods that meet the preset conditions in the matching logical trip set, α _p represents the overlapping duration corresponding to the pth first logical trip that does not meet the preset conditions in the matching logical trip set, P represents the total number of first logical trips that do not meet the preset conditions in the matching logical trip set,
represents the nth first logical trip in the first trip trajectory,
represents the shortest trip duration of the nth first logical trip, N represents the total number of the first logical trips in the first trip trajectory,
represents the qth second logical trip in the second logical trip that does not match each of the first logical trips of the first trip trajectory,
represents the shortest trip duration of the qth second logical trip, and Q represents the total number of second logical trips that do not match each first logical trip of the first trip trajectory.
根据权利要求1至3任一项所述的识别方法，其中，所述第一逻辑出行和所述第二逻辑出行均还包括构成一次逻辑出行的出行最短时长；The identification method according to any one of claims 1 to 3, wherein both the first logical trip and the second logical trip further include a shortest trip duration that constitutes a logical trip;

其中，所述根据所述衰减处理后的重叠时长计算出所述第一出行轨迹和所述第二出行轨迹的轨迹相似度，具体包括：Wherein, calculating the trajectory similarity between the first travel trajectory and the second travel trajectory according to the overlapping duration after the attenuation process specifically includes:

向所述匹配逻辑出行集中的所述第一逻辑出行的出行最短时长、与所述多个第二逻辑出行都不匹配的所述第一逻辑出行的出行最短时长、与所述多个第一逻辑出行都不匹配的所述第二逻辑出行的出行最短时长分别赋予不同的权重系数；To the shortest travel duration of the first logical trip in the matching logical trip set, the shortest trip duration of the first logical trip that does not match the plurality of second logical trips, and the shortest trip duration of the first logical trip that does not match the plurality of first logical trips. The shortest trip durations of the second logical trips that do not match the logical trips are respectively assigned different weighting coefficients;

根据所述衰减处理后的重叠时长以及被分别赋予不同权重系数的所述匹配逻辑出行集中的所述第一逻辑出行的出行最短时长、与所述多个第二逻辑出行都不匹配的所述第一逻辑出行的出行最短时长、与所述多个第一逻辑出行都不匹配的所述第二逻辑出行的出行最短时长，计算出所述第一出行轨迹和所述第二出行轨迹的轨迹相似度。According to the overlapping duration after the attenuation process and the shortest trip duration of the first logical trip in the matching logical trip set assigned different weighting coefficients, the trip duration that does not match the plurality of second logical trips The shortest trip duration of the first logical trip, the shortest trip duration of the second logical trip that does not match the multiple first logical trips, and the traces of the first trip trajectory and the second trip trajectory are calculated similarity.
根据权利要求6所述的识别方法，其中，所述根据所述衰减处理后的重叠时长以及被分别赋予不同权重系数的所述匹配逻辑出行集中的所述第一逻辑出行的出行最短时长、与所述多个第二逻辑出行都不匹配的所述第一逻辑出行的出行最短时长、与所述多个第一逻辑出行都不匹配的所述第二逻辑出行的出行最短时长，计算出所述第一出行轨迹和所述第二出行轨迹的轨迹相似度，具体包括：基于下面的式子3计算得到所述轨迹相似度，The identification method according to claim 6, wherein the overlapping duration after the attenuation process and the shortest travel duration of the first logical trip in the matching logical trip set assigned different weight coefficients respectively, and The shortest trip duration of the first logical trip that does not match the multiple second logical trips, and the shortest trip duration of the second logical trip that does not match the multiple first logical trips, calculate the The trajectory similarity of the first travel trajectory and the second travel trajectory specifically includes: calculating the trajectory similarity based on the following formula 3,

其中，TS _afc表示所述第一出行轨迹，TS _ap表示所述第二出行轨迹，Sim(TS _afc,TS _ap)表示所述第一出行轨迹和所述第二出行轨迹的轨迹相似度，M表示所述匹配逻辑出行集中的符合预设条件的不同出行时间段的总数量，α _p表示所述匹配逻辑出行集中的未符合预设条件的第p个第一逻辑出行对应的重叠时长，P表示所述匹配逻辑出行集中的未符合预设条件的第一逻辑出行的总数量，
表示所述匹配逻辑出行集中的第s个第一逻辑出行，
表示所述第s个第一逻辑出行的出行最短时长，S表示所述匹配逻辑出行集中的第一逻辑出行的总数量，
表示与第二出行轨迹的各个第二逻辑出行都不匹配的第一逻辑出行中的第r个第一逻辑出行，
表示所述第r个第一逻辑出行的出行最短时长，R表示与第二出行轨迹的各个第二逻辑出行都不匹配的第一逻辑出行中的第一逻辑出行的总数量，
表示与所述第一出行轨迹的各个第一逻辑出行都不匹配的第二逻辑出行中的第q个第二逻辑出行，
表示所述第q个第二逻辑出行的出行最短时长，Q表示与所述第一出行轨迹的各个第一逻辑出行都不匹配的第二逻辑出行中的第二逻辑出行的总数量，θ ₁、θ ₂和θ ₃分别表示不同的惩罚系数。 Wherein, TS _afc represents the first travel trajectory, TS _ap represents the second travel trajectory, Sim(TS _afc , TS _ap ) represents the trajectory similarity between the first travel trajectory and the second travel trajectory, M represents the total number of different travel time periods that meet the preset conditions in the matching logical trip set, α _p represents the overlapping duration corresponding to the pth first logical trip that does not meet the preset conditions in the matching logical trip set, P represents the total number of first logical trips that do not meet the preset conditions in the matching logical trip set,
represents the sth first logical trip in the matching logical trip set,
represents the shortest trip duration of the sth first logical trip, S represents the total number of first logical trips in the matching logical trip set,
represents the rth first logical trip among the first logical trips that do not match each of the second logical trips of the second trip trajectory,
represents the shortest trip duration of the rth first logical trip, R represents the total number of first logical trips in the first logical trips that do not match each second logical trip of the second trip trajectory,
represents the qth second logical trip in the second logical trip that does not match each of the first logical trips of the first trip trajectory,
represents the shortest trip duration of the qth second logical trip, Q represents the total number of second logical trips in the second logical trips that do not match the first logical trips of the first trip trajectory, θ ₁ , θ ₂ and θ ₃ represent different penalty coefficients, respectively.
根据权利要求1所述的识别方法，其中，所述第一逻辑出行和所述第二逻辑出行均还包括构成一次逻辑出行的进站站点、出站站点、进站时间和出站时间；The identification method according to claim 1, wherein each of the first logical trip and the second logical trip further comprises an inbound station, an outbound station, an inbound time and an outbound time that constitute a logical trip;

其中，所述对所述第一逻辑出行和所述第二逻辑出行进行匹配，以将与所述第二逻辑出行匹配的所述第一逻辑出行归类于匹配逻辑出行集，具体包括：The matching of the first logical trip and the second logical trip to classify the first logical trip matching the second logical trip into a matching logical trip set specifically includes:

利用所述第一逻辑出行的进站时间减去所述第二逻辑出行的进站时间以得到进站时长，以及利用所述第二逻辑出行的出站时间减去所述第一逻辑出行的出站时间以得到出站时长；Subtracting the inbound time of the second logical trip from the inbound time of the first logical trip to obtain the inbound duration, and subtracting the inbound time of the first logical trip from the outbound time of the second logical trip Outbound time to get outbound time;

获取从所述第一逻辑出行的进站站点到所述第二逻辑出行的进站站点之间的最短进站时长，以及获取从所述第二逻辑出行的出站站点到所述第一逻辑出行的出站站点之间的最短出站时长；Obtaining the shortest inbound duration from the inbound site of the first logical trip to the inbound site of the second logical trip, and acquiring from the outbound site of the second logical trip to the first logical trip The minimum outbound duration between outbound stops of the trip;

若所述进站时长和所述最短进站时长之和的绝对值小于一预设时长阈值，且所述出站时长和所述最短出站时长之和的绝对值小于所述预设时长阈值，则确定所述第一逻辑出行和所述第二逻辑出行匹配，并将与所述第二逻辑出行匹配的所述第一逻辑出行归类于匹配逻辑出行集。If the absolute value of the sum of the inbound duration and the shortest inbound duration is less than a preset duration threshold, and the absolute value of the sum of the outbound duration and the shortest outbound duration is less than the preset duration threshold , then it is determined that the first logical trip matches the second logical trip, and the first logical trip matching the second logical trip is classified into a matching logical trip set.
一种公共交通乘客的识别装置，其中，所述识别装置包括：An identification device for public transport passengers, wherein the identification device comprises:

出行轨迹获取模块，被配置为基于第一乘客出行数据源获取包括多个第一逻辑出行的第一出行轨迹，且基于第二乘客出行数据源获取包括多个第二逻辑出行的第二出行轨迹；其中，所述第一逻辑出行和所述第二逻辑出行分别包括构成一次逻辑出行的出行时间段；a travel trajectory obtaining module configured to obtain a first travel trajectory including a plurality of first logical trips based on the first passenger travel data source, and obtain a second travel trajectory including a plurality of second logical travels based on the second passenger travel data source ; wherein, the first logical trip and the second logical trip respectively include a trip time period that constitutes a logical trip;

逻辑出行匹配模块，被配置为对所述第一逻辑出行和所述第二逻辑出行进行匹配，以将与所述第二逻辑出行匹配的所述第一逻辑出行归类于匹配逻辑出行集；a logical trip matching module configured to match the first logical trip and the second logical trip to classify the first logical trip matching the second logical trip into a set of matching logical trips;

重叠时长获取模块，被配置为获取所述匹配逻辑出行集中的所述第一逻辑出行的出行时间段和其匹配的所述第二逻辑出行的出行时间段重叠的出行时间段，以得到重叠的出行时间段所对应的重叠时长；The overlapping duration acquisition module is configured to acquire the travel time period in which the travel time period of the first logical trip in the matching logical trip set and the travel time period of the matching second logical trip overlap, so as to obtain the overlapping travel time period. The overlapping duration corresponding to the travel time period;

重叠时长衰减模块，被配置为对所述重叠时长进行衰减处理，以得到衰减处理后的重叠时长；an overlapping duration attenuation module, configured to perform attenuation processing on the overlapping duration to obtain an overlapping duration after attenuation processing;

轨迹相似度计算模块，被配置为根据所述衰减处理后的重叠时长计算出所述第一出行轨迹和所述第二出行轨迹的轨迹相似度；a trajectory similarity calculation module, configured to calculate the trajectory similarity of the first travel trajectory and the second travel trajectory according to the overlapping duration after the attenuation process;

乘客识别模块，被配置为根据所述轨迹相似度来对第一乘客和第二乘客进行识别。The passenger identification module is configured to identify the first passenger and the second passenger according to the track similarity.
根据权利要求9所述的识别装置，其中，所述重叠时长衰减模块包括：The identification device according to claim 9, wherein the overlapping duration decay module comprises:

时长衰减单元，被配置为对所述匹配逻辑出行集中符合预设条件的所述第一逻辑出行所对应的所述重叠时长进行衰减，以得到衰减重叠时长；a duration attenuation unit, configured to attenuate the overlapping duration corresponding to the first logical trip that meets the preset condition in the matching logical trip set, so as to obtain the attenuation overlapping duration;

时长确定单元，被配置为将所述衰减重叠时长和所述匹配逻辑出行集中不符合所述预设条件的所述第一逻辑出行所对应的所述重叠时长确定为所述衰减处理后的重叠时长；A duration determination unit configured to determine the attenuation overlap duration and the overlap duration corresponding to the first logical trip that does not meet the preset condition in the matching logical trip set as the attenuation processed overlap duration;

其中，所述预设条件包括：在所述匹配逻辑出行集中的一个出行时间段中，对应该出行时间段的所有所述第一逻辑出行的出行时长之和大于预设时长。Wherein, the preset condition includes: in a travel time period in the matching logical travel set, the sum of travel time lengths of all the first logical travels corresponding to the travel time period is greater than a preset time period.
根据权利要求10所述的识别装置，其中，所述时长衰减单元进一步被配置为利用下面的式子1计算得到所述衰减重叠时长，The identification device according to claim 10, wherein the duration attenuation unit is further configured to obtain the attenuation overlap duration by calculating the following formula 1,

其中，a表示所述匹配逻辑出行集中的符合所述预设条件的第a个出行时间段，c _a表示所述第a个出行时间段中的所述第一逻辑出行的总数量，b表示第a个出行时间段中的第b个第一逻辑出行，γ表示衰减贡献率，
表示所述第b个第一逻辑出行所对应的重叠时长，
表示所述第b个第一逻辑出行所对应的衰减重叠时长。 Among them, a represents the a-th travel time period that meets the preset condition in the matching logical travel set, c _a represents the total number of the first logical trips in the a-th travel time period, and b represents The bth first logical trip in the ath trip time period, γ represents the decay contribution rate,
represents the overlapping duration corresponding to the bth first logical trip,
Indicates the decay overlap duration corresponding to the bth first logical row.
根据权利要求9至11任一项所述的识别装置，其中，所述第一逻辑出行和所述第二逻辑出行均还包括构成一次逻辑出行的出行最短时长；The identification device according to any one of claims 9 to 11, wherein each of the first logical trip and the second logical trip further includes a shortest trip duration constituting one logical trip;

所述轨迹相似度计算模块进一步被配置为：根据所述衰减处理后的重叠时长、所述多个第一逻辑出行的出行最短时长以及与所述多个第一逻辑出行都不匹配的所述第二逻辑出行的出行最短时长，计算出所述第一出行轨迹和所述第二出行轨迹的相似度。The trajectory similarity calculation module is further configured to: according to the overlapping duration after the attenuation process, the shortest travel duration of the plurality of first logical trips, and the said plurality of first logical trips that do not match. The shortest trip duration of the second logical trip is calculated, and the similarity between the first trip trajectory and the second trip trajectory is calculated.
根据权利要求12所述的识别装置，其中，所述轨迹相似度计算模块进一步被配置为利用下面的式子2计算得到所述轨迹相似度，The identification device according to claim 12, wherein the trajectory similarity calculation module is further configured to calculate the trajectory similarity using the following formula 2,

其中，TS _afc表示所述第一出行轨迹，TS _ap表示所述第二出行轨迹， Sim(TS _afc,TS _ap)表示所述第一出行轨迹和所述第二出行轨迹的轨迹相似度，M表示所述匹配逻辑出行集中的符合预设条件的不同出行时间段的数量，α _p表示所述匹配逻辑出行集中的未符合预设条件的第p个所述第一逻辑出行对应的重叠时长，P表示所述匹配逻辑出行集中的不符合预设条件的第一逻辑出行的总数量，
表示所述第一出行轨迹中的第n个第一逻辑出行，
表示所述第n个第一逻辑出行的出行最短时长，N表示所述第一出行轨迹中的所述第一逻辑出行的总数量，
表示与所述第一出行轨迹的各个第一逻辑出行都不匹配的第二逻辑出行中的第q个第二逻辑出行，
表示所述第q个第二逻辑出行的出行最短时长，Q表示与所述第一出行轨迹的各个第一逻辑出行都不匹配的第二逻辑出行的总数量。 Wherein, TS _afc represents the first travel trajectory, TS _ap represents the second travel trajectory, Sim(TS _afc , TS _ap ) represents the trajectory similarity between the first travel trajectory and the second travel trajectory, M represents the number of different travel time periods that meet the preset conditions in the matching logical trip set, α _p represents the overlapping duration corresponding to the pth first logical trip that does not meet the preset conditions in the matching logical trip set, P represents the total number of first logical trips that do not meet the preset conditions in the matching logical trip set,
represents the nth first logical trip in the first trip trajectory,
represents the shortest trip duration of the nth first logical trip, N represents the total number of the first logical trips in the first trip trajectory,
represents the qth second logical trip in the second logical trip that does not match each of the first logical trips of the first trip trajectory,
represents the shortest trip duration of the qth second logical trip, and Q represents the total number of second logical trips that do not match each first logical trip of the first trip trajectory.
根据权利要求9至11任一项所述的识别装置，其中，所述第一逻辑出行和所述第二逻辑出行均还包括构成一次逻辑出行的出行最短时长；The identification device according to any one of claims 9 to 11, wherein each of the first logical trip and the second logical trip further includes a shortest trip duration constituting one logical trip;

其中，所述轨迹相似度计算模块包括：Wherein, the trajectory similarity calculation module includes:

权重系数赋予单元，被配置为向所述匹配逻辑出行集中的所述第一逻辑出行的出行最短时长、与所述多个第二逻辑出行都不匹配的所述第一逻辑出行的出行最短时长、与所述多个第一逻辑出行都不匹配的所述第二逻辑出行的出行最短时长分别赋予不同的权重系数；A weighting coefficient assigning unit configured to assign the shortest travel duration of the first logical trip in the matching logical trip set and the shortest trip duration of the first logical trip that does not match the plurality of second logical trips , assigning different weight coefficients to the shortest travel durations of the second logical trips that do not match the multiple first logical trips;

相似度计算单元，被配置为根据所述衰减处理后的重叠时长以及被分别赋予不同权重系数的所述匹配逻辑出行集中的所述第一逻辑出行的出行最短时长、与所述多个第二逻辑出行都不匹配的所述第一逻辑出行的出行最短时长、与所述多个第一逻辑出行都不匹配的所述第二逻辑出行的出行最短时长，计算出所述第一出行轨迹和所述第二出行轨迹的轨迹相似度。The similarity calculation unit is configured to, according to the overlapping duration after the attenuation process and the shortest trip duration of the first logical trip in the matching logical trip set assigned different weight coefficients, and the plurality of second logical trips respectively. The shortest trip duration of the first logical trip that does not match the logical trips, and the shortest trip duration of the second logical trips that do not match the multiple first logical trips, calculate the first trip trajectory and The track similarity of the second travel track.
根据权利要求14所述的识别装置，其中，所述相似度计算单元进一步被配置为利用下面的式子3计算得到所述轨迹相似度，The identification device according to claim 14, wherein the similarity calculation unit is further configured to calculate the trajectory similarity using the following formula 3,

其中，TS _afc表示所述第一出行轨迹，TS _ap表示所述第二出行轨迹，Sim(TS _afc,TS _ap)表示所述第一出行轨迹和所述第二出行轨迹的轨迹相似度，M表示所述匹配逻辑出行集中的符合预设条件的不同出行时间段的总数量，α _p表示所述匹配逻辑出行集中的未符合预设条件的第p个第一逻辑出行对应的重叠时长，P表示所述匹配逻辑出行集中的未符合预设条件的第一逻辑出行的总数量，
表示所述匹配逻辑出行集中的第s个第一逻辑出行，
表示所述第s个第一逻辑出行的出行最短时长，S表示所述匹配逻辑出行集中的第一逻辑出行的总数量，
表示与第二出行轨迹的各个第二逻辑出行都不匹配的第一逻辑出行中的第r个第一逻辑出行，
表示所述第r个第一逻辑出行的出行最短时长，R表示与第二出行轨迹的各个第二逻辑出行都不匹配的第一逻辑出行中的第一逻辑出行的总数量，
表示与所述第一出行轨迹的各个第一逻辑出行都不匹配的第二逻辑出行中的第q个第二逻辑出行，
表示所述第q个第二逻辑出行的出行最短时长，Q表示与所述第一出行轨迹的各个第一逻辑出行都不匹配的第二逻辑出行中的第二逻辑出行的总数量，θ ₁、θ ₂和θ ₃分别表示不同的惩罚系数。 Wherein, TS _afc represents the first travel trajectory, TS _ap represents the second travel trajectory, Sim(TS _afc , TS _ap ) represents the trajectory similarity between the first travel trajectory and the second travel trajectory, M represents the total number of different travel time periods that meet the preset conditions in the matching logical trip set, α _p represents the overlapping duration corresponding to the pth first logical trip that does not meet the preset conditions in the matching logical trip set, P represents the total number of first logical trips that do not meet the preset conditions in the matching logical trip set,
represents the sth first logical trip in the matching logical trip set,
represents the shortest trip duration of the sth first logical trip, S represents the total number of first logical trips in the matching logical trip set,
represents the rth first logical trip among the first logical trips that do not match each of the second logical trips of the second trip trajectory,
represents the shortest trip duration of the rth first logical trip, R represents the total number of first logical trips in the first logical trips that do not match each second logical trip of the second trip trajectory,
represents the qth second logical trip in the second logical trip that does not match each of the first logical trips of the first trip trajectory,
represents the shortest trip duration of the qth second logical trip, Q represents the total number of second logical trips in the second logical trips that do not match the first logical trips of the first trip trajectory, θ ₁ , θ ₂ and θ ₃ represent different penalty coefficients, respectively.
根据权利要求9所述的识别装置，其中，所述第一逻辑出行和所述第二逻辑出行均还包括构成一次逻辑出行的进站站点、出站站点、进站时间和出站时间；The identification device according to claim 9, wherein each of the first logical trip and the second logical trip further comprises an inbound station, an outbound station, an inbound time and an outbound time that constitute a logical trip;

所述逻辑出行匹配模块包括：The logical trip matching module includes:

进出站时长确定单元，被配置为利用所述第一逻辑出行的进站时间减去所述第二逻辑出行的进站时间以得到进站时长，以及利用所述第二逻辑出行的出站时间减去所述第一逻辑出行的出站时间以得到出站时长；an inbound and outbound duration determination unit configured to subtract the inbound time of the second logical trip from the inbound time of the first logical trip to obtain the inbound duration, and the outbound time of the second logical trip Subtract the outbound time of the first logical trip to obtain the outbound duration;

最短进出站时长确定单元，被配置为获取从所述第一逻辑出行的进站站点到所述第二逻辑出行的进站站点之间的最短进站时长，以及获取从所述第二逻辑出行的出站站点到所述第一逻辑出行的出站站点之间的最短出站时长；The shortest inbound and outbound duration determination unit is configured to obtain the shortest inbound duration from the inbound station of the first logical trip to the inbound station of the second logical trip, and to acquire the distance from the second logical trip to the inbound station of the second logical trip The shortest outbound duration between the outbound site of the first logical trip and the outbound site of the first logical trip;

匹配确定单元，被配置为若所述进站时长和所述最短进站时长之和的绝对值小于一预设时长阈值，且所述出站时长和所述最短出站时长之和的绝对值小于所述预设时长阈值，则确定所述第一逻辑出行和所述第二逻辑出行匹配，并将与所述第二逻辑出行匹配的所述第一逻辑出行归类于匹配逻辑出行集。The matching determination unit is configured to, if the absolute value of the sum of the inbound duration and the shortest inbound duration is less than a preset duration threshold, and the absolute value of the sum of the outbound duration and the shortest outbound duration If it is less than the preset duration threshold, it is determined that the first logical trip matches the second logical trip, and the first logical trip matching the second logical trip is classified into a matching logical trip set.
一种电子设备，其中，包括：An electronic device comprising:

至少一个处理器，以及at least one processor, and

与所述至少一个处理器耦合的存储器，所述存储器存储指令，当所述指令被所述至少一个处理器执行时，使得所述至少一个处理器执行如权利要求1至8中任一所述的公共交通乘客的识别方法。a memory coupled to the at least one processor, the memory storing instructions that, when executed by the at least one processor, cause the at least one processor to perform the performance of any one of claims 1 to 8 method for identification of public transport passengers.
一种机器可读存储介质，其存储有可执行指令，其中，所述指令当被执行时使得所述机器执行如权利要求1到8中任一所述的公共交通乘客的识别方法。A machine-readable storage medium having executable instructions stored thereon, wherein the instructions, when executed, cause the machine to perform the method of identifying a public transportation passenger as claimed in any one of claims 1 to 8.