CN111178628A - Luggage arrival time prediction method and device - Google Patents
Luggage arrival time prediction method and device Download PDFInfo
- Publication number
- CN111178628A CN111178628A CN201911395473.5A CN201911395473A CN111178628A CN 111178628 A CN111178628 A CN 111178628A CN 201911395473 A CN201911395473 A CN 201911395473A CN 111178628 A CN111178628 A CN 111178628A
- Authority
- CN
- China
- Prior art keywords
- flight
- information
- sample
- baggage
- time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 43
- 238000000605 extraction Methods 0.000 claims abstract description 15
- 238000012549 training Methods 0.000 claims description 35
- 238000003860 storage Methods 0.000 claims description 3
- 230000002123 temporal effect Effects 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 8
- 230000008569 process Effects 0.000 description 12
- 230000006870 function Effects 0.000 description 11
- 230000004913 activation Effects 0.000 description 7
- 230000010006 flight Effects 0.000 description 6
- 239000013598 vector Substances 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000005192 partition Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 230000001413 cellular effect Effects 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000008034 disappearance Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000015654 memory Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/40—Business processes related to the transportation industry
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Human Resources & Organizations (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Economics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Strategic Management (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Quality & Reliability (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method and a device for predicting the arrival time of luggage, which are characterized in that the method comprises the steps of collecting the associated information of a flight where target luggage is located, wherein the associated information at least comprises the data of the flight and the preorder flight information corresponding to the flight; predicting to obtain the predicted time information of the target baggage reaching the target baggage carousel according to the associated information of the flight and a pre-established time prediction model; and pushing the predicted time information to a target passenger corresponding to the target luggage. According to the invention, a time prediction model can be obtained by combining and modeling according to the current flight information and the preorder flight information through multi-dimensional data, so that the time prediction model has larger learning information amount and stronger representation capability, and the time prediction result is more accurate. Therefore, the residence time of passengers at baggage extraction points is reduced, and the passenger experience effect is improved.
Description
Technical Field
The invention relates to the technical field of information processing, in particular to a luggage arrival time prediction method and device applied to a civil aviation airport.
Background
With the development of the aviation field, more and more passengers choose to take airplanes for traveling. When passengers take planes, they usually carry personal luggage and check in the luggage before boarding. After the passenger arrives at the destination airport, the passenger may arrive at the terminal building before the baggage, which may be transported to the designated baggage carousel via a uniform consignment process. The existing airport passengers usually inquire the position of a baggage carousel and the number of the baggage carousel corresponding to the flight on which the passengers take the airport intelligently, and cannot obtain the time for the baggage to reach the baggage carousel, so that the passengers can only stay near the baggage carousel for a long time to wait. The residence time of the passengers at the luggage extraction points is increased, so that the experience effect of the passengers is poor.
Disclosure of Invention
In view of the above problems, the present invention provides a baggage arrival time prediction method and device, which reduce the residence time of a passenger at a baggage extraction point and improve the experience effect of the passenger.
In order to achieve the purpose, the invention provides the following technical scheme:
a baggage arrival time prediction method, the method comprising
Acquiring the associated information of the flight where the target luggage is located, wherein the associated information at least comprises the data of the flight and the preorder flight information corresponding to the flight;
predicting to obtain the predicted time information of the target baggage reaching the target baggage carousel according to the associated information of the flight and a pre-established time prediction model;
and pushing the predicted time information to a target passenger corresponding to the target luggage.
Optionally, the method further comprises:
obtaining sample information, wherein the sample information represents data information related to the arrival time of the luggage in all current flight arrival time periods;
determining a label of each sample of the sample information, wherein the label is the time of arrival of a row matched with each sample to a luggage turntable;
and performing model training according to the sample information and the label of each sample of the sample information to obtain a time prediction model.
Optionally, the determining the label of each sample point of the sample information includes:
and acquiring the time of the baggage corresponding to each sample reaching the baggage tray through the radio frequency identification device.
Optionally, the obtaining sample information includes:
acquiring characteristic data of a current flight;
acquiring preamble flight data on the same time sequence with the current flight;
generating a single piece of sample information according to the characteristic data of the current flight and the preorder aviation data corresponding to the current flight;
and combining the plurality of pieces of sample information to obtain sample information.
Optionally, the characteristic data of the current flight includes one or more of the following data:
the flight departure time characteristic, the actual arrival time period of the flight, the arrival time period of the flight plan, the delay time of the flight, the model of the flight plane, the number of baggage carried by the flight plane, the arrival weather characteristic of the flight, the arrival temperature characteristic of the flight, the stop position number of the flight plane, the number of a baggage rotating disc corresponding to the flight, and the distance information between the stop position of the flight and the baggage rotating disc.
Optionally, the performing model training according to the sample information and the label of each sample of the sample information to obtain a time prediction model includes:
extracting the characteristics of the sample information, and performing characteristic representation on the extracted characteristics to obtain target characteristic information;
and training the target characteristic information and the label of each sample through an LSTM model to obtain a time prediction model.
An apparatus for predicting arrival time of baggage, the apparatus comprising:
the information acquisition unit is used for acquiring the associated information of the flight where the target luggage is located, wherein the associated information at least comprises the data of the flight and the preorder flight information corresponding to the flight;
the prediction unit is used for predicting and obtaining the predicted time information of the target luggage reaching the target luggage turntable according to the associated information of the flight and a pre-established time prediction model;
and the pushing unit is used for pushing the predicted time information to the target passenger corresponding to the target luggage.
Optionally, the apparatus further comprises:
the system comprises a sample acquisition unit, a storage unit and a control unit, wherein the sample acquisition unit is used for acquiring sample information, and the sample information represents data information related to the arrival time of the luggage in all current flight arrival time periods;
the label determining unit is used for determining a label of each sample of the sample information, wherein the label is the time when the row matched with each sample arrives at the luggage turntable;
and the model training unit is used for carrying out model training according to the sample information and the label of each sample of the sample information to obtain a time prediction model.
Optionally, the tag determination unit includes:
collecting the time of the luggage corresponding to each sample reaching a luggage loading tray through a radio frequency identification device;
the sample acquiring unit includes:
the first acquiring subunit is used for acquiring the characteristic data of the current flight;
the second acquisition subunit is used for acquiring the preamble flight data of the current flight in the same time sequence;
the generating subunit is used for generating single piece of sample information according to the characteristic data of the current flight and the preorder aviation data corresponding to the current flight;
the combination subunit is used for combining the plurality of pieces of sample information to obtain sample information;
the characteristic data of the current flight comprises one or more of the following data:
the flight departure time characteristic, the actual arrival time period of the flight, the arrival time period of the flight plan, the delay time of the flight, the model of the flight plane, the number of baggage carried by the flight plane, the arrival weather characteristic of the flight, the arrival temperature characteristic of the flight, the stop position number of the flight plane, the number of a baggage rotating disc corresponding to the flight, and the distance information between the stop position of the flight and the baggage rotating disc.
Optionally, the model training unit comprises:
the characteristic extraction subunit is used for carrying out characteristic extraction on the sample information and carrying out characteristic representation on the extracted characteristics to obtain target characteristic information;
and the training subunit is used for training the target characteristic information and the label of each sample through an LSTM model to obtain a time prediction model.
Compared with the prior art, the invention provides a method and a device for predicting the arrival time of luggage, wherein the method comprises the steps of collecting the associated information of the flight where the target luggage is located, wherein the associated information at least comprises the data of the flight and the preorder flight information corresponding to the flight; predicting to obtain the predicted time information of the target baggage reaching the target baggage carousel according to the associated information of the flight and a pre-established time prediction model; and pushing the predicted time information to a target passenger corresponding to the target luggage. According to the invention, a time prediction model can be obtained by combining and modeling according to the current flight information and the preorder flight information through multi-dimensional data, so that the time prediction model has larger learning information amount and stronger representation capability, and the time prediction result is more accurate. Therefore, the residence time of passengers at baggage extraction points is reduced, and the passenger experience effect is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
Fig. 1 is a schematic flowchart of a baggage arrival time prediction method according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of data slicing according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of a baggage arrival time prediction apparatus according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The terms "first" and "second," and the like in the description and claims of the present invention and the above-described drawings are used for distinguishing between different objects and not for describing a particular order. Furthermore, the terms "comprising" and "having," as well as any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not set forth for a listed step or element but may include steps or elements not listed.
The invention provides a baggage arrival time prediction method, which is applied to civil aviation airports, wherein baggage arrival refers to arrival of baggage at a baggage tray or a designated baggage picking point. Referring to fig. 1, the method may include:
s101, collecting the associated information of the flight where the target luggage is located.
The associated information at least comprises the flight data and the preamble flight information corresponding to the flight. The information collected in the embodiment of the application not only includes data information of a flight where the current target baggage is located, but also includes M previous flight sets which may affect the arrival time of the current flight baggage in the time sequence.
S102, predicting to obtain the predicted time information of the target baggage reaching the target baggage carousel according to the associated information of the flight and a pre-established time prediction model.
The time prediction model is pre-created through sample information and has the function of predicting the time when the target baggage reaches the target baggage carousel. Correspondingly, the embodiment of the present application further provides a model creating method, and the method further includes:
obtaining sample information, wherein the sample information represents data information related to the arrival time of the luggage in all current flight arrival time periods;
determining a label of each sample of the sample information, wherein the label is the time of arrival of a row matched with each sample to a luggage turntable;
and performing model training according to the sample information and the label of each sample of the sample information to obtain a time prediction model.
Specifically, the time when the baggage corresponding to each sample arrives at the baggage tray may be collected by the rfid device. The process of obtaining the sample information is as follows: acquiring characteristic data of a current flight; acquiring preamble flight data on the same time sequence with the current flight; generating a single piece of sample information according to the characteristic data of the current flight and the preorder aviation data corresponding to the current flight; and combining the plurality of pieces of sample information to obtain sample information. The characteristic data of the current flight comprises one or more of the following data: the flight departure time characteristic, the actual arrival time period of the flight, the arrival time period of the flight plan, the delay time of the flight, the model of the flight plane, the number of baggage carried by the flight plane, the arrival weather characteristic of the flight, the arrival temperature characteristic of the flight, the stop position number of the flight plane, the number of a baggage rotating disc corresponding to the flight, and the distance information between the stop position of the flight and the baggage rotating disc.
Specifically, the acquisition of training data in the training process acquires all flight related information (including all data information related to the arrival time of baggage in all current flight arrival time periods) all day by using a real-time civil aviation service system. And recording the time when the luggage carried by the specific flight on the day arrives at the luggage turntable through the RFID-based hardware, and taking the time as result prediction data in the model training process. And constructing a time prediction model based on all the preorder flight data of the current flight and all the dimensional data corresponding to the current flight. And (3) using an LSTM algorithm, using an influence factor set of the current flight to the baggage arrival time, combining the influence factor sets of the current M preceding flights to serve as single training data, and using the predicted arrival time of the current flight baggage as label data to perform model training. And predicting the arrival time of the current flight baggage by using the current flight related data and combining all the previous flight data.
S103, pushing the predicted time information to a target passenger corresponding to the target luggage.
For example, the baggage arrival time of the current flight is pushed to the traveler in advance through various modes such as an electronic screen, a short message service (sms), an app message and the like.
The invention provides a luggage arrival time prediction method, which comprises the steps of collecting the associated information of a flight where a target luggage is located, wherein the associated information at least comprises the data of the flight and the preorder flight information corresponding to the flight; predicting to obtain the predicted time information of the target baggage reaching the target baggage carousel according to the associated information of the flight and a pre-established time prediction model; and pushing the predicted time information to a target passenger corresponding to the target luggage. According to the invention, a time prediction model can be obtained by combining and modeling according to the current flight information and the preorder flight information through multi-dimensional data, so that the time prediction model has larger learning information amount and stronger representation capability, and the time prediction result is more accurate. Therefore, the residence time of passengers at baggage extraction points is reduced, and the passenger experience effect is improved.
The temporal prediction model created in the present application is specifically described below.
First, characteristic information affecting the arrival time of the baggage needs to be determined. The flight baggage arrival time measuring method mainly comprises two dimensions, wherein one dimension is characteristic data related to a current flight, and the other dimension is a set of M previous flights possibly influencing the baggage arrival time of the current flight in time series.
The current flight related characteristic data comprises: whether the time is holiday (that day), flight actual arrival time period, planned arrival time period, flight delay time, carrier aircraft model number, carrier aircraft luggage number, aircraft arrival weather (that day), aircraft arrival temperature (current time period), carrier aircraft stop number, luggage extraction point turntable number, and distance between stop and luggage turntable.
The preamble flight related feature data includes: the current day is distinguished according to the baggage carousel number, and all flight information sets and all flight related data feature sets of the current day that are served by each baggage carousel are described in the above description of feature data related to the current flight (which is not described herein again).
For example: the flight number CA1234 [ whether it is a holiday (that day), a flight actual arrival time period, a planned arrival time period, a flight delay time, a carrier aircraft model number, a carrier aircraft baggage number, an aircraft arrival time weather (that day), an aircraft arrival time temperature (that current period), a carrier aircraft parking lot number, a baggage pickup spot carousel number, a distance between a parking lot and a baggage carousel ], a flight number CA6181 [ whether it is a holiday, a flight actual arrival time, an expected arrival time, a carrier aircraft model number, a carrier aircraft baggage number, an aircraft arrival time weather, an aircraft arrival time temperature, an aircraft parking lot number, a baggage pickup carousel number, a parking lot to baggage carousel distance ]. And counting M strips.
And a single training data generation rule takes the baggage carousel as a main key to acquire the relevant characteristic data of all historical flights of the current baggage carousel.
Assuming that the current baggage carousel generates related baggage feature records of N flights in total, starting from the first piece of data, the data is sliced by using M as the data slicing window size, and the N flight records are subjected to moving slicing by using the step length t. Referring to fig. 2, a schematic diagram of data slicing provided by the embodiment of the present application is shown. By using the method, all historical data are segmented to generate training data. Each box in fig. 2 represents a flight data (characteristic data relating to the current flight that may have an effect on the arrival time of the baggage) arriving at the current baggage carousel.
And a luggage partition board is used, the current luggage partition board is used for representing the flight number identification loaded by the current conveyor belt and reminding a user to take luggage, and meanwhile, an RFID electronic tag is arranged in the luggage partition board. The exit of the baggage carousel is provided with an RFID card reader for recording the arrival time of the baggage of a specific flight (arrival at the entrance of the baggage carousel) when all baggage carousels output. The method is used for predicting a target predicted value for model training.
The derived variables are constructed as variable factors that may have an effect on the arrival time of the baggage.
And if the holiday is a festival, presetting the holiday of the whole year. For example, the day uses 0 to represent not holidays, and 1 to represent that the day is holidays, and the holiday information of the day can be weighted, such as 0.4 for ordinary weekends, 0.7 for eleven, and 0.8 for monthly and weekends.
Flight actual arrival time period: the actual arrival time of the flight is mapped to three hours, morning, daytime and night. Morning (5: 01-9: 00), daytime (9: 01-20: 00), night (20: 01-5: 00).
The flight plan arrival time period is the same as the "flight actual arrival time period".
And (3) flight delay time, namely the actual arrival time of the airplane-the planned arrival time of the airplane (unit: min). The result is a value type, and the value type is used for unified processing.
And carrying out coded representation by using One-Hot type vectors when the airplane model is carried.
The baggage number of the airplane is carried, and the baggage consignment system can acquire the information of the current baggage number of the airplane. The representation is transformed using a numerical type variable.
And weather when the airplane arrives, and the part of data can be obtained through a weather forecast query interface. Including "sunny", "light rain", "medium rain", "heavy rain", "small snow", "medium snow", "heavy snow", "typhoon", etc. The model of the carrier airplane is expressed by using a One-Hot type vector form.
The temperature of the aircraft at arrival, the data source, is as above, and is represented using a numerical type variable transform.
And carrying the number of the airplane parking space, wherein an airport data system can obtain the number and the number is coded and expressed by using One-Hot type vectors.
The luggage extraction point turntable numbers are obtained by a luggage consignment system, and are coded and expressed by using One-Hot type vectors.
The distance between the parking space and the luggage rotary table and the position of the luggage rotary table are represented by the entrance position of the luggage van of the airport terminal. Namely, the actual positions of the parking positions and the luggage turnplate are represented by the distance between the parking positions and freight ports of each station building, the partial data can be directly inquired and acquired in a knowledge map of airport information, and the data source is data which is sorted and recorded when indoor navigation is made in the prior airport. A numerical type transformation is used.
And (3) carrying out normalization processing by using min-max on the numerical type variable transformation expression: scaling the range of values to (0,1) without changing the data distribution; this algorithm is a linear transformation of the original data, with the result falling in the [0,1] interval, and the transfer function is as follows:
x=(x-min)/(max-min)
max is the maximum value of the sample data, min is the minimum value of the sample data.
An empirical constant value is used in place of max and min in the patent numerical variable conversion process. Using historical 30-day homogeneous data, the two boundary values within the 95% confidence interval were used as the maximum and minimum values, respectively. All the same-class data of 30 days are taken out, the mean value and the variance are obtained, and the data range can be obtained by using the mean value, the variance and the 95% confidence interval.
The One-Hot basic idea is that the One-Hot type vector is used for coding representation: each value of the discrete feature is considered as a state, if N different values exist in the feature, the feature can be abstracted into N different states, and one-hot coding ensures that each value only enables one state to be in an activated state, namely only one state bit value in the N states is 1, and other state bits are 0. For example, all airplane models of an airport are 5(D) in total, including 717, 737, 747, 757 and 767, and can be identified by using 5(D) dimensional data, such as 717 represented by [10000] and 737 represented by [010000 ].
According to the method and the device for predicting the arrival time of the passenger baggage at the baggage carousel, the arrival time of the passenger baggage can be influenced by the arrival time of the baggage of a preceding flight in addition to various characteristic factors of the environment where the current flight is located, and the whole process is sequentially executed in the time dimension, and the prediction of the arrival time of the baggage of a certain flight in the process can be understood as a typical prediction process based on a time sequence. The model is predicted using the LSTM algorithm.
The LSTM model structure includes 1 input layer (input dimension is determined by flight related baggage feature data dimension), 1 fully connected output layer.
The loss function (loss function) uses Mean Square Error (MSE), the optimization algorithm (Optimizer) uses adam, the batch size (batch _ size) is set to 64, the learning rate is set to 0.001, Dropout is set to 0.2, and the model weights are updated using a stochastic gradient descent method. And after the hyper-parameter optimization, applying the model to the test set, and if the predicted deviation value is within the required range, storing the model.
And acquiring data of the current flight after being processed by the characteristic engineering and the processed data of the previous M flights of the current flight, inputting the data into a prediction model, and predicting the time of the baggage of the current flight reaching a baggage carousel of a baggage extraction point.
In the embodiment of the application, the time sequence prediction is carried out by using the LSTM deep learning framework, compared with the traditional machine learning algorithm, the prediction accuracy is higher, the robustness and the fault-tolerant capability on noise nerves are stronger, and the dynamic neural network LSTM is more suitable for predicting the time sequence than a static neural network BP network. Various feature information related to the current flight and historical feature information of the preorder flight are applied to combined modeling, the information amount used in the modeling process is larger, and the model has stronger representation capability. Using lstm and setting the step size to 1, the number of training data samples is greatly increased. The time windows of M preceding flights are used for manufacturing test data, on one hand, the number of training samples is increased, on the other hand, the training efficiency is improved, the training of the whole data in one day is avoided, the time is not divided according to the day, and the data representation in the time dimension is continuous (for example, the baggage arrival time of the flight at 1 point 10 minutes in the morning can be predicted by using the data in a certain time window in the previous day). By using supervised learning, the real data after the system is on-line can be used for continuously enriching the training data sample base, continuously adjusting and optimizing the existing model, and improving the model prediction capability. The tag data is acquired, the luggage segmentation card with the RFID is used for extracting, the part of data is uniformly recorded and accurately identified through the system, the problem of shielding (image identification of the digital card is carried out firstly), the identification effective space is large, the maintenance cost is low during trial, the whole process is completed automatically, and the manual intervention is minimized.
The LSTM deep learning framework is explained below.
LSTM changes the memory in a very precise way, applying specialized learning mechanisms to remember, update, and focus on the information. This helps to track information over a longer period of time. In a Recurrent Neural Network (RNN) model and a forward back propagation algorithm, as the RNN also has the problem of gradient disappearance, Long-sequence data is difficult to process, the RNN is improved, and a special case LSTM (Long Short-Term Memory) of the RNN is obtained, so that the gradient disappearance of a conventional RNN can be avoided. The internal structure of the LSTM model at each sequence index position t is as follows: in addition to the hidden state h (t) h (t) which is the same as RNN, there is another hidden state that propagates forward at each sequence index position t. This hidden state we generally refer to as the cell state (CellState), denoted C (t). In addition to the cellular state, LSTM is also called a gated structure (Gate). The gates of the LSTM at each sequence index position t typically include three types, a forgetting gate, an input gate, and an output gate.
The forgetting gate, the input gate, the output gate, and the cell state of the LSTM will be specifically described below.
As the name implies, the forgetting gate (forget gate) controls whether to forget, and in LSTM, controls whether to forget the state of the hidden cell in the previous layer with a certain probability. The forgetting gate substructure input has a previous sequence of hidden states h (t-1) h (t-1) and the sequence data x (t) x (t), and the output f (t) f (t) of the forgetting gate is obtained through an activation function, generally sigmoid. Since the output f (t) f (t) of sigmoid is between [0,1], the output f { (t) } represents the probability of forgetting the state of the previous layer of hidden cells. The mathematical expression is as follows:
f(t)=σ(Wfh(t-1)+Ufx(t)+bf)f(t)=σ(Wfh(t-1)+Ufx(t)+bf)
where Wf, Uf, bfWf, Uf, bf are coefficients and biases of linear relationships, similar to those in RNN. σ σ is the sigmoid activation function.
The input gate (input gate) is responsible for processing the input of the current sequence position, and consists of two parts, wherein the first part uses a sigmoid activation function, the output is i (t) i (t), the second part uses a tanh activation function, the output is a (t) a (t), and the results of the two parts are multiplied later to update the cell state. The mathematical expression is as follows:
i(t)=σ(Wih(t-1)+Uix(t)+bi)i(t)=σ(Wih(t-1)+Uix(t)+bi)
a(t)=tanh(Wah(t-1)+Uax(t)+ba)a(t)=tanh(Wah(t-1)+Uax(t)+ba)
where Wi, Ui, bi, Wa, Ua, ba, coefficients and biases are linear relationships, similar to those in RNN. σ σ is the sigmoid activation function.
Cellular status of LSTM. The results of the forgoing forgetting gate and the entry gate both contribute to the cell state C (t). How to obtain C (t) C (t) from the cell state C (t-1) C (t-1) is described below.
The cell state C (t) consists of two parts, the first part being the product of C (t-1) and the forgetting gate output f (t) f (t), and the second part being the product of i (t) i (t) and a (t) a (t) of the input gate, i.e.:
wherein C (t) ═ C (t-1) ⊙ f (t) + i (t) ⊙ a) (C (t-1) ⊙ f (t) + i (t) ⊙ a (t) >, as a Hadamard product, as well as in DNN.
The outputs are as follows:
the updating of the hidden state h (t) h (t) consists of two parts, the first part is o (t) o (t), which is obtained by the previous sequence of hidden states h (t-1) h (t-1) and the sequence data x (t) x (t), and the activation function sigmoid, and the second part consists of the hidden states C (t) and the tanh activation function, namely:
o(t)=σ(Woh(t-1)+Uox(t)+bo)o(t)=σ(Woh(t-1)+Uox(t)+bo)
h(t)=o(t)⊙tanh(C(t))h(t)=o(t)⊙tanh(C(t))
in an embodiment of the present application, there is provided a baggage arrival time prediction apparatus, referring to fig. 3, the apparatus including:
the information acquisition unit 10 is configured to acquire associated information of a flight on which the target baggage is located, where the associated information at least includes data of the flight and flight preamble information corresponding to the flight;
the prediction unit 20 is configured to predict and obtain predicted time information when the target baggage reaches the target baggage carousel according to the associated information of the flight and a pre-established time prediction model;
a pushing unit 30, configured to push the predicted time information to a target passenger corresponding to the target baggage.
On the basis of the above embodiment, the apparatus further includes:
the system comprises a sample acquisition unit, a storage unit and a control unit, wherein the sample acquisition unit is used for acquiring sample information, and the sample information represents data information related to the arrival time of the luggage in all current flight arrival time periods;
the label determining unit is used for determining a label of each sample of the sample information, wherein the label is the time when the row matched with each sample arrives at the luggage turntable;
and the model training unit is used for carrying out model training according to the sample information and the label of each sample of the sample information to obtain a time prediction model.
On the basis of the above embodiment, the tag determination unit includes:
collecting the time of the luggage corresponding to each sample reaching a luggage loading tray through a radio frequency identification device;
the sample acquiring unit includes:
the first acquiring subunit is used for acquiring the characteristic data of the current flight;
the second acquisition subunit is used for acquiring the preamble flight data of the current flight in the same time sequence;
the generating subunit is used for generating single piece of sample information according to the characteristic data of the current flight and the preorder aviation data corresponding to the current flight;
the combination subunit is used for combining the plurality of pieces of sample information to obtain sample information;
the characteristic data of the current flight comprises one or more of the following data:
the flight departure time characteristic, the actual arrival time period of the flight, the arrival time period of the flight plan, the delay time of the flight, the model of the flight plane, the number of baggage carried by the flight plane, the arrival weather characteristic of the flight, the arrival temperature characteristic of the flight, the stop position number of the flight plane, the number of a baggage rotating disc corresponding to the flight, and the distance information between the stop position of the flight and the baggage rotating disc.
On the basis of the above embodiment, the model training unit includes:
the characteristic extraction subunit is used for carrying out characteristic extraction on the sample information and carrying out characteristic representation on the extracted characteristics to obtain target characteristic information;
and the training subunit is used for training the target characteristic information and the label of each sample through an LSTM model to obtain a time prediction model.
The invention provides a baggage arrival time prediction device, which is characterized in that the correlation information of a flight where a target baggage is located is acquired, wherein the correlation information at least comprises the data of the flight and the preorder flight information corresponding to the flight; predicting to obtain the predicted time information of the target baggage reaching the target baggage carousel according to the associated information of the flight and a pre-established time prediction model; and pushing the predicted time information to a target passenger corresponding to the target luggage. According to the invention, a time prediction model can be obtained by combining and modeling according to the current flight information and the preorder flight information through multi-dimensional data, so that the time prediction model has larger learning information amount and stronger representation capability, and the time prediction result is more accurate. Therefore, the residence time of passengers at baggage extraction points is reduced, and the passenger experience effect is improved.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A baggage arrival time prediction method, characterized in that the method comprises:
acquiring the associated information of the flight where the target luggage is located, wherein the associated information at least comprises the data of the flight and the preorder flight information corresponding to the flight;
predicting to obtain the predicted time information of the target baggage reaching the target baggage carousel according to the associated information of the flight and a pre-established time prediction model;
and pushing the predicted time information to a target passenger corresponding to the target luggage.
2. The method of claim 1, further comprising:
obtaining sample information, wherein the sample information represents data information related to the arrival time of the luggage in all current flight arrival time periods;
determining a label of each sample of the sample information, wherein the label is the time of arrival of a row matched with each sample to a luggage turntable;
and performing model training according to the sample information and the label of each sample of the sample information to obtain a time prediction model.
3. The method of claim 2, wherein the determining the label for each sample point of the sample information comprises:
and acquiring the time of the baggage corresponding to each sample reaching the baggage tray through the radio frequency identification device.
4. The method of claim 2, wherein obtaining sample information comprises:
acquiring characteristic data of a current flight;
acquiring preamble flight data on the same time sequence with the current flight;
generating a single piece of sample information according to the characteristic data of the current flight and the preorder aviation data corresponding to the current flight;
and combining the plurality of pieces of sample information to obtain sample information.
5. The method of claim 4, wherein the current flight profile comprises one or more of the following:
the flight departure time characteristic, the actual arrival time period of the flight, the arrival time period of the flight plan, the delay time of the flight, the model of the flight plane, the number of baggage carried by the flight plane, the arrival weather characteristic of the flight, the arrival temperature characteristic of the flight, the stop position number of the flight plane, the number of a baggage rotating disc corresponding to the flight, and the distance information between the stop position of the flight and the baggage rotating disc.
6. The method of claim 5, wherein the model training according to the sample information and the label of each sample of the sample information to obtain a temporal prediction model comprises:
extracting the characteristics of the sample information, and performing characteristic representation on the extracted characteristics to obtain target characteristic information;
and training the target characteristic information and the label of each sample through an LSTM model to obtain a time prediction model.
7. An apparatus for predicting arrival time of baggage, the apparatus comprising:
the information acquisition unit is used for acquiring the associated information of the flight where the target luggage is located, wherein the associated information at least comprises the data of the flight and the preorder flight information corresponding to the flight;
the prediction unit is used for predicting and obtaining the predicted time information of the target luggage reaching the target luggage turntable according to the associated information of the flight and a pre-established time prediction model;
and the pushing unit is used for pushing the predicted time information to the target passenger corresponding to the target luggage.
8. The apparatus of claim 7, further comprising:
the system comprises a sample acquisition unit, a storage unit and a control unit, wherein the sample acquisition unit is used for acquiring sample information, and the sample information represents data information related to the arrival time of the luggage in all current flight arrival time periods;
the label determining unit is used for determining a label of each sample of the sample information, wherein the label is the time when the row matched with each sample arrives at the luggage turntable;
and the model training unit is used for carrying out model training according to the sample information and the label of each sample of the sample information to obtain a time prediction model.
9. The apparatus according to claim 8, wherein the tag determination unit comprises:
collecting the time of the luggage corresponding to each sample reaching a luggage loading tray through a radio frequency identification device;
the sample acquiring unit includes:
the first acquiring subunit is used for acquiring the characteristic data of the current flight;
the second acquisition subunit is used for acquiring the preamble flight data of the current flight in the same time sequence;
the generating subunit is used for generating single piece of sample information according to the characteristic data of the current flight and the preorder aviation data corresponding to the current flight;
the combination subunit is used for combining the plurality of pieces of sample information to obtain sample information;
the characteristic data of the current flight comprises one or more of the following data:
the flight departure time characteristic, the actual arrival time period of the flight, the arrival time period of the flight plan, the delay time of the flight, the model of the flight plane, the number of baggage carried by the flight plane, the arrival weather characteristic of the flight, the arrival temperature characteristic of the flight, the stop position number of the flight plane, the number of a baggage rotating disc corresponding to the flight, and the distance information between the stop position of the flight and the baggage rotating disc.
10. The apparatus of claim 9, wherein the model training unit comprises:
the characteristic extraction subunit is used for carrying out characteristic extraction on the sample information and carrying out characteristic representation on the extracted characteristics to obtain target characteristic information;
and the training subunit is used for training the target characteristic information and the label of each sample through an LSTM model to obtain a time prediction model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911395473.5A CN111178628B (en) | 2019-12-30 | 2019-12-30 | Luggage arrival time prediction method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911395473.5A CN111178628B (en) | 2019-12-30 | 2019-12-30 | Luggage arrival time prediction method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111178628A true CN111178628A (en) | 2020-05-19 |
CN111178628B CN111178628B (en) | 2023-09-19 |
Family
ID=70654298
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911395473.5A Active CN111178628B (en) | 2019-12-30 | 2019-12-30 | Luggage arrival time prediction method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111178628B (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112396254A (en) * | 2020-12-11 | 2021-02-23 | 北京百分点信息科技有限公司 | Destination prediction method, destination prediction device, destination prediction medium, and electronic device |
CN112836905A (en) * | 2021-04-22 | 2021-05-25 | 中航信移动科技有限公司 | Flight event occurrence time prediction method, device, equipment and storage medium |
CN113449915A (en) * | 2021-06-28 | 2021-09-28 | 中国电子科技集团公司第二十八研究所 | Flight delay prediction method based on knowledge graph |
CN113835331A (en) * | 2021-09-10 | 2021-12-24 | 西安悦泰科技有限责任公司 | Method for acquiring and displaying passenger baggage extraction pre-waiting time |
CN116596168A (en) * | 2023-07-17 | 2023-08-15 | 四川省机场集团有限公司 | Airport luggage information synchronization and prediction method and system |
CN116703010A (en) * | 2023-08-09 | 2023-09-05 | 民航成都信息技术有限公司 | Luggage arrival time prediction method and device, electronic equipment and medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2004046522A (en) * | 2002-07-11 | 2004-02-12 | Mitsubishi Electric Corp | Aircraft arrival information delivery system, processing server and terminal |
CN106778857A (en) * | 2016-12-09 | 2017-05-31 | 北京首都国际机场股份有限公司 | To the method that port flight baggage status are obtained automatically |
CN108197081A (en) * | 2017-11-03 | 2018-06-22 | 上海敬之网络科技有限公司 | A kind of data actuarial model method for building up of flight delay danger |
CN108388963A (en) * | 2018-02-13 | 2018-08-10 | 百度在线网络技术(北京)有限公司 | Luggage arrival time prediction technique, device, computer equipment and readable medium |
CN109448445A (en) * | 2018-09-05 | 2019-03-08 | 南京航空航天大学 | Flight based on shot and long term Memory Neural Networks is delayed grading forewarning system method |
CN109791731A (en) * | 2017-06-22 | 2019-05-21 | 北京嘀嘀无限科技发展有限公司 | A kind of method and system for estimating arrival time |
-
2019
- 2019-12-30 CN CN201911395473.5A patent/CN111178628B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2004046522A (en) * | 2002-07-11 | 2004-02-12 | Mitsubishi Electric Corp | Aircraft arrival information delivery system, processing server and terminal |
CN106778857A (en) * | 2016-12-09 | 2017-05-31 | 北京首都国际机场股份有限公司 | To the method that port flight baggage status are obtained automatically |
CN109791731A (en) * | 2017-06-22 | 2019-05-21 | 北京嘀嘀无限科技发展有限公司 | A kind of method and system for estimating arrival time |
CN108197081A (en) * | 2017-11-03 | 2018-06-22 | 上海敬之网络科技有限公司 | A kind of data actuarial model method for building up of flight delay danger |
CN108388963A (en) * | 2018-02-13 | 2018-08-10 | 百度在线网络技术(北京)有限公司 | Luggage arrival time prediction technique, device, computer equipment and readable medium |
CN109448445A (en) * | 2018-09-05 | 2019-03-08 | 南京航空航天大学 | Flight based on shot and long term Memory Neural Networks is delayed grading forewarning system method |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112396254A (en) * | 2020-12-11 | 2021-02-23 | 北京百分点信息科技有限公司 | Destination prediction method, destination prediction device, destination prediction medium, and electronic device |
CN112836905A (en) * | 2021-04-22 | 2021-05-25 | 中航信移动科技有限公司 | Flight event occurrence time prediction method, device, equipment and storage medium |
CN113449915A (en) * | 2021-06-28 | 2021-09-28 | 中国电子科技集团公司第二十八研究所 | Flight delay prediction method based on knowledge graph |
CN113449915B (en) * | 2021-06-28 | 2023-10-27 | 中国电子科技集团公司第二十八研究所 | Flight delay prediction method based on knowledge graph |
CN113835331A (en) * | 2021-09-10 | 2021-12-24 | 西安悦泰科技有限责任公司 | Method for acquiring and displaying passenger baggage extraction pre-waiting time |
CN116596168A (en) * | 2023-07-17 | 2023-08-15 | 四川省机场集团有限公司 | Airport luggage information synchronization and prediction method and system |
CN116703010A (en) * | 2023-08-09 | 2023-09-05 | 民航成都信息技术有限公司 | Luggage arrival time prediction method and device, electronic equipment and medium |
CN116703010B (en) * | 2023-08-09 | 2023-10-31 | 民航成都信息技术有限公司 | Luggage arrival time prediction method and device, electronic equipment and medium |
Also Published As
Publication number | Publication date |
---|---|
CN111178628B (en) | 2023-09-19 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111178628B (en) | Luggage arrival time prediction method and device | |
Manna et al. | A statistical approach to predict flight delay using gradient boosted decision tree | |
CN110555990B (en) | Effective parking space-time resource prediction method based on LSTM neural network | |
CN111401601B (en) | Delay propagation-oriented flight take-off and landing time prediction method | |
Chen et al. | Uncertainty in urban mobility: Predicting waiting times for shared bicycles and parking lots | |
Ballakur et al. | Empirical evaluation of gated recurrent neural network architectures in aviation delay prediction | |
CN115953186B (en) | Network appointment vehicle demand pattern recognition and short-time demand prediction method | |
CN109086926B (en) | Short-time rail transit passenger flow prediction method based on combined neural network structure | |
Dhief et al. | Predicting aircraft landing time in extended-TMA using machine learning methods | |
CN111784049B (en) | Passenger loss time prediction method and device | |
CN113657671B (en) | Flight delay prediction method based on ensemble learning | |
CN110443448B (en) | Bidirectional LSTM-based airplane position classification prediction method and system | |
CN111242395B (en) | Method and device for constructing prediction model for OD (origin-destination) data | |
CN112651534B (en) | Method, device and storage medium for predicting resource supply chain demand | |
CN112132366A (en) | Prediction system for flight clearance rate | |
CN115358491A (en) | Subway passenger flow prediction method and system based on CEEMDAN-LSTM | |
Zhang et al. | An interpretable station delay prediction model based on graph community neural network and time-series fuzzy decision tree | |
CN110796315A (en) | Departure flight delay prediction method based on aging information and deep learning | |
Ariyawansa et al. | Review on state of art data mining and machine learning techniques for intelligent Airport systems | |
CN116432806A (en) | Rolling prediction method and system for flight ground guarantee node time | |
CN113221472B (en) | Passenger flow prediction method based on LSTM | |
Han et al. | Short-term 4D trajectory prediction based on LSTM neural network | |
Hatıpoğlu et al. | Flight delay prediction based with machine learning | |
Faridai et al. | Neural network model for predicting passenger congestion to optimize traffic management for urban public transport | |
Ayaydın et al. | Deep learning based forecasting of delay on flights |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |