CN116205329A - Holiday passenger flow prediction method - Google Patents
Holiday passenger flow prediction method Download PDFInfo
- Publication number
- CN116205329A CN116205329A CN202211601798.6A CN202211601798A CN116205329A CN 116205329 A CN116205329 A CN 116205329A CN 202211601798 A CN202211601798 A CN 202211601798A CN 116205329 A CN116205329 A CN 116205329A
- Authority
- CN
- China
- Prior art keywords
- holiday
- trend
- passenger flow
- item
- term
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 18
- 230000000737 periodic effect Effects 0.000 claims abstract description 19
- 230000001932 seasonal effect Effects 0.000 claims abstract description 18
- 238000004140 cleaning Methods 0.000 claims abstract description 5
- 238000012549 training Methods 0.000 claims abstract description 4
- 238000012886 linear function Methods 0.000 claims description 8
- 238000007477 logistic regression Methods 0.000 claims description 7
- 230000003442 weekly effect Effects 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 7
- 238000005457 optimization Methods 0.000 abstract 1
- 230000006870 function Effects 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000007 visual effect Effects 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- 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
- G06Q50/10—Services
- G06Q50/14—Travel agencies
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Tourism & Hospitality (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Databases & Information Systems (AREA)
- Marketing (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a holiday passenger flow volume prediction method, which comprises the following steps: (1) statistics of holiday passenger flow historical data of recent three years; (2) cleaning historical data; (3) Based on a prophet algorithm, constructing a time sequence model of non-periodic variation trend item, seasonal trend item, holiday trend item, distance trend item, epidemic trend item, error item and weight item occupied by each trend item: (4) Substituting the cleaned historical data into each time sequence model for training; and (5) predicting the passenger flow based on the trained model. According to the method, optimization is performed on the basis of a traditional prophet algorithm, epidemic factors and distance factors are increased, different weights are determined for different trend items, an optimized prophet++ model is obtained, the real situation can be better fitted, and a prediction result with a better effect and accuracy is obtained.
Description
Technical Field
The invention relates to a holiday passenger flow prediction method, and belongs to the technical field of big data.
Background
Along with the development of national economy, the living standard of people is continuously improved, the consumption concept is also greatly changed, and the number and quality requirements of people in travel are also greatly changed. Particularly, a great deal of travel demands are generated during the holidays of the first year, the Qingming day, the fifth year, the end noon, the mid-autumn day, the tenth festival and the like, so that peak passenger flow is formed. In the peak stage of passenger traffic, the passenger flow is greatly increased, and only a scientific and flexible transportation organization scheme is established, the congestion condition of holidays can be effectively reduced. The scientific decision is not separated from scientific prediction, so that the traffic flow of holidays is predicted in advance, the traffic capacity and the vigor are reasonably arranged, and the congestion condition can be effectively reduced. However, most of the current holiday passenger flow prediction algorithms adopt a unitary linear regression algorithm, and the algorithm simply fits the trend according to the past passenger flow data, ignores the influence of seasonal holiday effects and the current social large environment, and is not accurate enough in predicted data.
Disclosure of Invention
The invention aims to solve the technical problem of providing a holiday passenger flow volume prediction method which can better fit the actual situation and enable the passenger flow volume prediction result to be more accurate.
In order to solve the technical problems, the technical scheme of the invention is as follows:
a holiday passenger flow volume prediction method comprises the following steps:
(1) Counting historical data of passenger flow in holidays of the recent three years;
(2) Cleaning historical data, and improving data quality;
(3) Based on a prophet algorithm, constructing a time sequence model of non-periodic variation trend item, seasonal trend item, holiday trend item, distance trend item, epidemic trend item, error item and weight item occupied by each trend item:
W(t)＝(α 1 g(t)+α 2 s(t)+α 3 h(t)+α 4 d(t)*y(t)+e(t)
wherein W (t) represents an overall trend model, g (t) represents a non-periodic variation trend term, s (t) represents a seasonal trend term, h (t) represents a holiday trend term, d (t) represents a distance trend term, y (t) represents an epidemic trend term, e (t) represents an error term, and α 1 Representing the weight of the non-periodic variation trend term, alpha 2 Representing the weight occupied by seasonal trend term, alpha 3 Representing the weight occupied by holiday trend term, alpha 4 Representing the weight occupied by the distance trend item;
(4) Substituting the cleaned historical data into each time sequence model for training;
(5) And predicting the passenger flow based on the trained model.
As a preferable solution, in the step (1), the holiday traffic history data of the recent three years may be counted according to different holiday travel modes, where the holiday travel modes include: urban traffic, rail traffic, aviation, passenger transport and high-speed rail cars.
As a preferable scheme, the data cleaning in the step (2) includes clearing empty values and abnormal values in the data so as to improve the quality of the historical data.
As a preferred solution, the non-periodic variation trend term g (t) in the step (3) includes a trend term based on logistic regression and a trend term based on piecewise linear function, where the trend term based on logistic regression is:
c is the bearing capacity, k is the growth rate, and m is the offset parameter; wherein c=c (t), k=k (t);
trend term based on piecewise linear function:
k represents the rate of increase, δ represents the amount of change in the rate of increase, and m represents the amount of offset.
As a preferred scheme, the seasonal trend term s (t) in the step (3):
p represents the period and the parameters form a column vector, which can be expressed as:
β＝[a 1 ，b 1 ，...，a N ，b N ] T
for a one year periodic sequence (p=365.25), n=10; for a weekly-periodic sequence (p=7), n=3,
when n=10:
when n=3:
the last seasonal term s (t) =x (t) β, where β -Normal (0, σ) 2 )。
As a preferable scheme, the holiday trend term h (t) in the step (3) is:
κ～Normal(0，v 2 )，D i representing the time period before and after the holiday, κ i The influence range of holidays is represented, and L represents the number of holidays.
As a preferable mode, the distance trend term d (t) in the step (3):
As a preferred scheme, the epidemic trend term y (t) in the step (3):
x is the difference between the number of people in the last year and the current year of the corresponding time period, and the value range of y (t) is (1, 2).
As a preferable mode, α in the step (3) 1 Take the value of 0.5, alpha 2 Take the value of 0.2, alpha 3 Take the value of 0.2, alpha 4 The value is 0.1.
The invention has the beneficial effects that: compared with the prior art, the method optimizes the prophet algorithm, increases epidemic factors and distance factors, simultaneously determines different weights for different trend items to obtain an optimized prophet++ model, and can better fit the real situation to obtain a predicted result with better effect. According to the invention, through analysis and research on holiday travel behavior characteristics, holiday traffic characteristics can be understood more deeply, and an important theoretical basis is provided for holiday traffic planning and demand management.
The invention has the following characteristics: (1) The traditional prophet algorithm and the current epidemic factor influence factor are combined, so that the holiday traffic prediction of the passenger festival achieves a better effect; (2) The tidal characteristics of local trip, peripheral trip and short trip taking the regional center city as the center of the circle are combined, so that the accuracy of passenger flow prediction is greatly enhanced; (3) Different weights are given to different trend items, so that the trend items are more in line with actual conditions, and the prediction result is more accurate.
Currently affected by epidemic situation, holiday passenger flow mainly presents the following two characteristics: firstly, the passenger flow volume is obviously reduced compared with the past year. The annual pseudo-period national passenger traffic volume is 1 hundred million times, 2000 ten thousand times per day, and is reduced by about 62 percent compared with the annual pseudo-period 2021. Secondly, the passenger flow travel is mainly 'on-site and nearby going-out'. The travel radius of the holiday masses is obviously reduced compared with the travel radius of the holiday masses in a normal period under the influence of epidemic situation and prevention and control policies in various places. Traditional passenger flow prediction is performed through personal experience and visual analysis capability, and has great uncertainty. Therefore, the algorithm of the invention adds the factors of seasons, holidays, distances, epidemic situations and the like, and can better fit the actual situation, so that the passenger flow prediction result is more accurate and is closer to the actual situation.
Drawings
FIG. 1 is a graph of the number of trips versus distance in a distance trend of the present invention.
The invention is further described below with reference to the drawings and the detailed description.
Detailed Description
Example 1: the holiday passenger flow prediction method of the invention comprises the following steps:
(1) Counting the historical data of the passenger flow of the holidays of the recent three years according to different holiday travel modes, wherein the holiday travel modes comprise: urban traffic, rail traffic, aviation, passenger transport and high-speed rail cars;
(2) Cleaning historical data, clearing empty values and abnormal values in the data, and improving the data quality;
(3) Based on a prophet algorithm, constructing a time sequence model of non-periodic variation trend item, seasonal trend item, holiday trend item, distance trend item, epidemic trend item, error item and weight item occupied by each trend item:
W(t)＝(α 1 g(t)+α 2 s(t)+α 3 h(t)+α 4 d(t))*y(t)+e(t)
wherein W (t) represents an overall trend model, and g (t) represents an aperiodic variation trend term which represents an aperiodic variation trend of the time sequence; s (t) represents a seasonal trend term that represents a periodic variation in time series (e.g., monthly and per monthSeasonal changes in year); h (t) represents holiday trend terms, reflecting the effect of holidays occurring on a possibly irregular schedule; d (t) represents a distance trend term which represents the influence of the distance between the region and the regional center city; y (t) represents epidemic trend items and represents the influence of the epidemic situation in the region; e (t) represents an error term representing a fluctuation not predicted by the model and assuming that it conforms to a normal distribution; alpha 1 Representing the weight of the non-periodic variation trend term, alpha 2 Representing the weight occupied by seasonal trend term, alpha 3 Representing the weight occupied by holiday trend term, alpha 4 Representing the weight occupied by the distance trend item; the individual items are fitted separately and the final accumulated result is the predicted result of the algorithm.
The propset algorithm (time series prediction algorithm) is a time series prediction algorithm which is an open source by Facebook team in 2017, belongs to the prior art, combines time series decomposition and machine learning algorithms, and can predict time series with missing values and abnormal values.
(4) Substituting the cleaned historical data into each time sequence model for training, and improving the accuracy of each model;
(5) And predicting the passenger flow based on the trained model.
The following describes each time series model in the present invention:
the non-periodic variation trend term g (t) comprises a trend term based on logistic regression and a trend term based on piecewise linear function, wherein the trend term based on logistic regression is as follows:
c is the bearing capacity, k is the growth rate, and m is the offset parameter;
in the internet era, the development of hardware and the upgrading of information quantity are not constant, and the bearing capacity and the growth rate are not constant, so that the model modifies the fixed bearing capacity and the growth rate into a function of the bearing capacity and the growth rate which change with time, wherein C=C (t) and k=k (t);
trend term based on piecewise linear function: in reality, there will be a variable point in the time sequence, that is, where the trend changes, when the variable point is set, the model changes to be a piecewise linear function model:
k represents the rate of increase, δ represents the amount of change in the rate of increase, and m represents the amount of offset.
The greatest difference between piecewise linear functions, in which γ= (γ) 1 ，...，γ S ) T ，γ j ＝s j δ j This is not the same as the setting in the prior logistic regression function.
(II) seasonal trend term s (t):
the passenger travel rate typically varies seasonally with seasonal changes in day, week, month, year, etc., also referred to as periodic changes. The invention establishes a period model through Fourier series (Fourier series):
p represents the period and the parameters form a column vector, which can be expressed as:
β＝[a 1 ，b 1 ，...，a N ，b N ] T
the adjustment of N acts as a low-pass filter. The applicant finds that the effect of selecting N as 10 and 3 is better for the annual and weekly periods, respectively, i.e. n=10 for a sequence of one year periods (p=365.25); for a weekly-periodic sequence (p=7), n=3,
when n=10:
when n=3:
the last seasonal term s (t) =x (t) β, where β -Normal (0, σ) 2 )。
(III) holiday trend term h (t)
The holiday has extremely great influence on passenger flow, and particularly, the people in the holiday are rich in national countries, convenient in traffic and increasingly more in number of people going out of the holiday. Because the holidays of each festival have different influence on the time sequence, for example, spring festival, national festival is a seven-day holiday, and holidays of labor festival and the like are shorter. Thus, different holidays can be viewed as models that are independent of each other, and different front and rear window values can be set for different holidays, indicating that the holiday affects the time series of the front and rear periods.
In mathematical language, for the ith holiday Di represents the time before and after the holiday, a corresponding indication function (Indicator function) is required to represent the holiday effect, while a parameter ki is required to represent the holiday's range of influence. Assuming that we have L holidays, then:
wherein:
Z(t)＝[1(t∈D 1 )，...，1(t∈D L )]
κ～Normal(0，v 2 )
k accords with normal distribution, the default value is 10, and when the value is larger, the influence of holidays on the model is larger; the smaller the value, the less effective the holiday is on the model. The user can adjust according to the situation.
(IV) distance trend term d (t)
The tidal characteristics of local trip, peripheral trip and short trip taking regional center city as the center appear in the current vacation public trip, the applicant counts the holiday passenger flow number of each region in certain province 2020-2022, finds that the number of the trips and the distance are in a two-stage curve, basically meets the eccentric lognormal distribution (as shown in figure 1), and can obtain the trend item of the distance and the passenger flow:
where X is the distance of the travel area from the center city, σ is the density situation, given by default as 0.5, a is a constant greater than 1, and the default is 10.
(V) epidemic trend item y (t)
In recent years, epidemic situation is huge for people going out on holidays, and the applicant calculates epidemic situation influence factors of the corresponding holidays according to the comparison of 2022 (data before holidays) and people in the corresponding stages of the first two years:
(VI) weight occupied by each trend item:
the traditional propset algorithm is consistent in trend weights such as holidays, seasonality and the like, but in actual conditions, the influence weights of factors such as non-periodic variation trend, holidays, seasons, distances and the like on passenger flows are different, the applicant tests 2015-2022 passenger flow data fitting of certain province, and finds that the non-periodic variation trend is the heaviest, then the holidays and the seasons and finally the distances, so that the occupied weight values of trend items are as follows: alpha 1 Is 0.5, alpha 2 Is 0.2, alpha 3 Is 0.2, alpha 4 0.1.
In order to test the accuracy of the algorithm of the invention, the applicant correspondingly predicts five passenger flows in 2022 on the basis of three years of historical data of 2019, 2020 and 2021 from the aspects of whole road network, urban roads, expressways, cemetery and scenic spots, and the like, and the result shows that the predicted hit rate in other aspects is above 85 percent except the expressways, railways and civil aviation passenger flows (complete data cannot be acquired), and the calculated comprehensive hit rate is 92.46 percent, and the highest hit rate is even 97.68 percent. The hit rate of the prediction of urban roads, scenic spots and business district congestion sections is 100%, so that the prediction of the present invention for the future congestion sections is well known, and the present invention can be used as a reference for public travel.
The embodiments of the present invention are not limited to the above examples, and various changes made without departing from the spirit of the present invention are all within the scope of the present invention.
Claims (9)
1. The holiday passenger flow volume prediction method is characterized by comprising the following steps of:
(1) Counting historical data of passenger flow in holidays of the recent three years;
(2) Cleaning the historical data;
(3) Based on a prophet algorithm, constructing a time sequence model of non-periodic variation trend item, seasonal trend item, holiday trend item, distance trend item, epidemic trend item, error item and weight item occupied by each trend item:
W(t)＝(α 1 g(t)+α 2 s(t)+α 3 h(t)+α 4 d(t))*y(t)+e(t)
wherein W (t) represents an overall trend model, g (t) represents a non-periodic variation trend term, st) represents a seasonal trend term, h (t) represents a holiday trend term, d (t) represents a distance trend term, y (t) represents an epidemic trend term, e (t) represents an error term, and α 1 Representing the weight of the non-periodic variation trend term, alpha 2 Representing the weight occupied by seasonal trend term, alpha 3 Representing the weight occupied by holiday trend term, alpha 4 Representing the weight occupied by the distance trend item;
(4) Substituting the cleaned historical data into each time sequence model for training;
(5) And predicting the passenger flow based on the trained model.
2. The holiday traffic prediction method according to claim 1, wherein in the step (1), the holiday traffic history data of the recent three years is counted according to different holiday travel modes, and the holiday travel modes include: urban traffic, rail traffic, aviation, passenger transport and high-speed rail cars.
3. The holiday passenger flow volume prediction method of claim 1, wherein the data cleansing in step (2) comprises clearing empty values and outliers in the data.
4. The holiday passenger flow volume prediction method according to claim 1, wherein the non-periodic variation trend term g (t) in the step (3) includes a trend term based on logistic regression and a trend term based on piecewise linear function, and the trend term based on logistic regression is:
c is the bearing capacity, k is the growth rate, and m is the offset parameter; wherein c=c (t), k=k (t);
trend term based on piecewise linear function:
k represents the rate of increase, δ represents the amount of change in the rate of increase, and m represents the amount of offset.
5. The holiday passenger flow volume prediction method according to claim 1, wherein the seasonal trend term s (t) in the step (3):
p represents the period and the parameters form a column vector, which can be expressed as:
for a one year periodic sequence, n=10; for a weekly periodic sequence, n=3,
when n=10:
when n=3:
the last seasonal term s (t) =x (t) β, where β -Normal (0, σ) 2 )。
6. The holiday passenger flow volume prediction method according to claim 1, wherein the holiday trend term h (t) in the step (3) is:
Z(t)＝[1(t∈D 1 )，...，1(t∈D L )]
κ～Normal(0，v 2 )，D i representing the time period before and after the holiday, κ i The influence range of holidays is represented, and L represents the number of holidays.
7. The holiday passenger flow volume prediction method according to claim 1, wherein the distance trend term d (t) in the step (3):
8. The holiday passenger flow volume prediction method according to claim 1, wherein the epidemic trend term y (t) in the step (3):
9. The holiday passenger flow volume prediction method according to claim 1, wherein α in the step (3) 1 Is 0.5, alpha 2 Is 0.2, alpha 3 Is 0.2, alpha 4 0.1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211601798.6A CN116205329A (en) | 2022-12-13 | 2022-12-13 | Holiday passenger flow prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211601798.6A CN116205329A (en) | 2022-12-13 | 2022-12-13 | Holiday passenger flow prediction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116205329A true CN116205329A (en) | 2023-06-02 |
Family
ID=86508433
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211601798.6A Pending CN116205329A (en) | 2022-12-13 | 2022-12-13 | Holiday passenger flow prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116205329A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117118907A (en) * | 2023-10-25 | 2023-11-24 | 深圳市亲邻科技有限公司 | Entrance guard flow dynamic monitoring system and method thereof |
| CN117474299A (en) * | 2023-12-27 | 2024-01-30 | 南京满鲜鲜冷链科技有限公司 | Prediction method, device and equipment for cold transport supply and demand |
-
2022
- 2022-12-13 CN CN202211601798.6A patent/CN116205329A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117118907A (en) * | 2023-10-25 | 2023-11-24 | 深圳市亲邻科技有限公司 | Entrance guard flow dynamic monitoring system and method thereof |
| CN117118907B (en) * | 2023-10-25 | 2024-02-02 | 深圳市亲邻科技有限公司 | Entrance guard flow dynamic monitoring system and method thereof |
| CN117474299A (en) * | 2023-12-27 | 2024-01-30 | 南京满鲜鲜冷链科技有限公司 | Prediction method, device and equipment for cold transport supply and demand |
| CN117474299B (en) * | 2023-12-27 | 2024-02-27 | 南京满鲜鲜冷链科技有限公司 | Prediction method, device and equipment for cold transport supply and demand |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108564790B (en) | Urban short-term traffic flow prediction method based on traffic flow space-time similarity | |
| CN116205329A (en) | Holiday passenger flow prediction method | |
| CN110555990B (en) | Effective parking space-time resource prediction method based on LSTM neural network | |
| CN109376935B (en) | Bus passenger flow time interval combined prediction method based on neural network | |
| CN109615226A (en) | A kind of operation indicator method for monitoring abnormality | |
| CN107194491A (en) | A kind of dynamic dispatching method based on Forecasting of Travel Time between bus passenger flow and station | |
| CN113723659B (en) | Urban rail transit full-scene passenger flow prediction method and system | |
| CN111860989B (en) | LSTM neural network short-time traffic flow prediction method based on ant colony optimization | |
| CN110956807A (en) | Highway flow prediction method based on combination of multi-source data and sliding window | |
| Wu et al. | Predicting peak load of bus routes with supply optimization and scaled Shepard interpolation: A newsvendor model | |
| Zhang et al. | Agent-based model of price competition, capacity choice, and product differentiation on congested networks | |
| CN109147324B (en) | Traffic jam probability forecasting method based on user feedback mechanism | |
| CN115359659B (en) | Lane opening and closing configuration method and system | |
| CN112036670A (en) | Passenger flow comprehensive prediction system and prediction method | |
| CN108805727A (en) | method and device for data processing | |
| CN107368915B (en) | Subway passenger travel time selection behavior analysis method | |
| CN112613637A (en) | Method and device for processing charging load | |
| CN114066503A (en) | Method for forecasting taxi demand based on virtual service area construction | |
| CN110020475A (en) | A kind of Markov particle filter method of forecasting traffic flow | |
| CN115953186A (en) | Network appointment demand pattern recognition and short-time demand prediction method | |
| CN114037138A (en) | Subway short-time arrival passenger flow prediction system based on double-layer decomposition and deep learning and implementation method | |
| CN114187766A (en) | Road service level evaluation method based on saturation rate | |
| CN110490365B (en) | Method for predicting network car booking order quantity based on multi-source data fusion | |
| CN110119884B (en) | High-speed railway passenger flow time interval division method based on neighbor propagation clustering | |
| CN113537596A (en) | Short-time passenger flow prediction method for new line station of urban rail transit |
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 |