CN114328675A - Bus travel time prediction method based on double-attention machine system and bidirectional double-layer LSTM - Google Patents
Bus travel time prediction method based on double-attention machine system and bidirectional double-layer LSTM Download PDFInfo
- Publication number
- CN114328675A CN114328675A CN202210007787.9A CN202210007787A CN114328675A CN 114328675 A CN114328675 A CN 114328675A CN 202210007787 A CN202210007787 A CN 202210007787A CN 114328675 A CN114328675 A CN 114328675A
- Authority
- CN
- China
- Prior art keywords
- bus
- travel time
- time
- double
- data set
- 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
Images
Landscapes
- Traffic Control Systems (AREA)
Abstract
The invention relates to a bus travel time prediction method based on a double-attention mechanism and a bidirectional double-layer LSTM, comprising the following steps of S1, acquiring bus attribute data and station entering and exiting data, and constructing a travel time basic data set; s2, acquiring basic characteristics, running time and weather characteristics of the bus during running, and constructing a travel time characteristic data set; step S3, based on correlation coefficient and variance analysis, carrying out correlation analysis on the characteristic factors in the travel time characteristic data set and the travel time, eliminating the characteristic factors which are irrelevant and have poor correlation, and matching with the travel time basic data set to obtain a travel time prediction data set; and S4, constructing a double-attention machine mechanism and a bidirectional double-layer LSTM neural network bus travel time prediction model, inputting the travel time prediction data set into the model, predicting the travel time of the bus, and outputting the predicted travel time of the bus. The invention can accurately predict the travel time of the bus and realize high-efficiency bus scheduling.
Description
Technical Field
The invention relates to the field of urban intelligent traffic, in particular to a bus travel time prediction method based on a double-attention machine mechanism and a bidirectional double-layer LSTM.
Background
The bus as an important component of public transportation has the advantages of small per capita area, low per capita energy consumption, large bearing capacity, low overall transportation cost and the like. Through the service level that promotes the bus, attract more people to use public transport, help alleviating urban traffic and block up, reduce urban air pollution, ensure healthy good operation in city. Therefore, the rapid development of the bus is an important measure for realizing the sustainable development of urban traffic, promoting the virtuous circle of people and society and realizing interactive development.
At present, the bus scheduling is mainly carried out by estimating from own experience by scheduling personnel, or the predicted travel time is obtained based on a traditional bus travel time prediction model (such as a historical data model, a time sequence model, a regression prediction model, a support vector machine model, a Kalman filtering model and the like). The estimation error of the method is often larger, the scheduling is not timely enough, the phenomenon of 'train crossing' or 'large interval' is caused, the operation efficiency of the bus is lower, the service level is poorer, the attraction is not high, and the utilization rate does not reach the ideal level.
Disclosure of Invention
In view of the above, the present invention aims to provide a method for predicting bus travel time based on a dual-attention mechanism and a bidirectional double-layer LSTM, which can more accurately predict the bus travel time and realize efficient bus scheduling.
In order to achieve the purpose, the invention adopts the following technical scheme:
a bus travel time prediction method based on a double-attention mechanism and a bidirectional double-layer LSTM comprises the following steps:
s1, acquiring bus attribute data and station entering and exiting data, preprocessing the data and constructing a travel time basic data set;
s2, acquiring basic characteristics, running time and weather characteristics of the bus during running, and constructing a travel time characteristic data set;
step S3, based on correlation coefficient and variance analysis, carrying out correlation analysis on the characteristic factors in the travel time characteristic data set and the travel time, eliminating the characteristic factors which are irrelevant and have poor correlation, and matching with the travel time basic data set to obtain a travel time prediction data set;
and S4, constructing a double-attention machine mechanism and a bidirectional double-layer LSTM neural network bus travel time prediction model, inputting the travel time prediction data set into the model, predicting the travel time of the bus, and outputting the predicted travel time of the bus.
Further, the step S1 is specifically:
s11, acquiring bus attribute data and station entering and exiting data, directly eliminating repeated records and abnormal data and complementing missing data;
and step S12, matching the attribute data of the bus with the station entering and exiting data according to the serial number of the bus operation shifts, acquiring the bus operation time of each operation shift, and constructing a travel time basic data set.
Further, the step S2 is specifically: acquiring a bus number, a driver number, bus departure time and a bus departure interval of a bus; acquiring the running time from the starting station to the terminal station in the similar time period of the previous day and the similar time period of the same characteristic day of the previous week; and acquiring weather characteristics including weather type, humidity, wind speed and temperature when the vehicle runs, and constructing a travel time characteristic data set.
Further, the step S3 is specifically:
step S31: performing correlation analysis between the continuous characteristic factors and the bus running time by using the correlation coefficient of the Pearson correlation coefficient to obtain the correlation between variables;
step S32: classifying and coding discrete characteristic factors such as weather types, driver numbers, bus numbers and the like, converting the discrete characteristic factors into classification characteristic factors, and analyzing the correlation between the classification characteristic factors and the bus running time by using variance test;
step S33: and (4) discarding the characteristic factors which are irrelevant and have poor correlation, and matching the retained characteristics with the travel time basic data set to obtain a travel time prediction data set.
Further, the step S4 is specifically:
step S41: inputting the travel time prediction data set into a feature importance extraction module, and distributing attention of different degrees to the features according to the importance difference of the features to obtain a feature importance matrix;
step S42: splicing the characteristic importance matrix with the input characteristics of the model to obtain a bus running state matrix;
step S43: inputting the bus running state matrix into a time characteristic extraction module, extracting time characteristics in the bus running process by using a bidirectional LSTM, and generating a running time characteristic matrix;
step S44: inputting the obtained running time characteristic matrix into a travel time prediction module, fusing Attention into the time step of the LSTM, and constructing an Attention _ LSTM layer to realize extraction of influence characteristics of buses with different time distances; in the full-connection layer, the root mean square error is used as a loss function, the prediction result is compared with the actual result, and the model is continuously trained; the prediction of the bus travel time is realized.
A bus travel time prediction system based on a double-attention mechanism and a bidirectional double-layer LSTM comprises
The feature importance extraction module is used for acquiring importance weights of all features by using an attention mechanism to generate a feature importance matrix, distinguishing the influence of different features on the travel time of the bus and improving the prediction efficiency and the prediction precision of the model;
the time characteristic extraction module is used for capturing the running characteristics of the buses running before and after by utilizing the stronger time characteristic capture capability of the bidirectional LSTM, and acquiring the time characteristics of the buses in the running process;
and the travel time prediction module introduces Attention into the LSTM, constructs an Attention _ LSTM layer, gives different Attention to different time steps in the LSTM and expresses the influence of vehicles at different time distances on the travel time of the bus.
Compared with the prior art, the invention has the following beneficial effects:
the invention introduces LSTM into the prediction of bus travel time, on one hand, bidirectional LSTM is used to capture the time characteristics of the bus before and after the bus runs; on the other hand, on the basis of the bidirectional LSTM, an Attention _ LSTM layer is added, so that the extraction of the influence characteristics of the buses with different time distances is realized, and the prediction precision of the model is effectively improved; the more accurate predicted travel time of the bus and the efficient bus dispatching are realized.
Drawings
FIG. 1 is a schematic flow diagram of the process of the present invention;
FIG. 2 is a diagram of a bus travel time prediction model architecture for a dual-attention mechanism and bi-directional dual-layer LSTM;
FIG. 3 is a B2 bus route diagram of the embodiment of the present invention;
FIG. 4 is a comparison of travel time predictions and actual values obtained in an embodiment of the present invention.
Detailed Description
The invention is further explained below with reference to the drawings and the embodiments.
Referring to fig. 1, the invention provides a bus travel time prediction method based on a double-attention mechanism and a bidirectional double-layer LSTM, comprising the following steps:
and step S1, acquiring the attribute data and the station entering and exiting data of the bus, directly eliminating repeated records and abnormal data and complementing missing data. Matching the attribute data of the bus with the station entering and exiting data according to the serial number of the bus operation classes, acquiring the bus operation time of each operation class, and constructing a travel time basic data set;
s2, acquiring basic characteristics of a bus, such as a bus number, a driver number, bus departure time, departure intervals and the like; acquiring the running time from the starting station to the terminal station in the similar time period of the previous day and the similar time period of the same characteristic day of the previous week; acquiring weather characteristics such as weather types, humidity, wind speeds and temperatures when a vehicle runs, and constructing a travel time characteristic data set;
step S3, based on correlation coefficient and variance analysis, carrying out correlation analysis on the characteristic factors in the characteristic data set and the travel time, discarding the characteristic factors which are irrelevant and have poor correlation, and matching the characteristic factors with the travel time basic data set to obtain a travel time prediction data set;
and S4, constructing a double-attention mechanism and bidirectional double-layer LSTM neural network bus travel time prediction model based on the attention mechanism and the LSTM, inputting the travel time prediction data set into the model, predicting the travel time of the bus, and outputting the predicted travel time of the bus.
Example 1:
referring to fig. 3, in the embodiment, B2 public transportation vehicles in guangzhou city are taken as research objects, B2 data of buses traveling on late at peak (17:00-19:00) in working days of 10/1/2020 to 12/11/2020 is selected as example data, and the specific implementation mode is as follows:
and step S1, acquiring the attribute data and the station entering and exiting data of the bus, directly eliminating repeated records and abnormal data and complementing missing data. And matching the attribute data of the bus with the station entering and exiting data according to the serial number of the bus operation shifts, acquiring the bus operation time of each operation shift, and constructing a travel time basic data set.
1. Data is collected. The experimental data collected included two parts: the first part is static data containing station base information on a B2 bus line; the second part is dynamic data, including B2 bus station entrance and exit data and bus attribute data.
The platform basic information comprises a platform number, platform longitude and latitude, platform ascending and descending identification and the like; the bus in-and-out data comprises a road list ID, a vehicle number, a platform name, an in-and-out time, an up-and-down identification, a date and the like; the bus attribute data includes a waybill ID, departure time, a vehicle number, a driver number, a route type, a departure interval, and the like.
2. And (4) preprocessing data. The data preprocessing mainly comprises the steps of bus station entering and exiting data preprocessing, bus attribute data preprocessing, static data and dynamic data matching and travel time obtaining.
The bus dispatching data preprocessing method mainly comprises the steps of removing repeatedly recorded data and extracting whole-course operation data.
The method aims at preprocessing the bus station entering and exiting data, and comprises the steps of removing repeated data, calibrating abnormal data, interpolating missing data and acquiring bus travel time. Firstly, directly eliminating repeatedly recorded station entering and exiting data; secondly, processing abnormal data, wherein the abnormal data refers to data with the conditions of time reversal, running time far less than the running time of an average site and the like, and the abnormal data is removed firstly and then is subjected to interpolation filling; and thirdly, processing the missing data, and mainly focusing on the data missing in the arrival time of the terminal station. Calculating the ratio of the average running time to the average stopping time of each part in the missing station, calculating the interpolation time of each part in the missing time period, completing, and finally acquiring the bus travel time of the running shift according to the departure time and the arrival time of the terminal station.
The static data and the dynamic data are matched. And matching the platform basic information, the bus attribute data and the bus in-and-out data according to the serial number of the bus operation class to obtain a travel time basic data set.
S2, acquiring basic characteristics of a bus, such as a bus number, a driver number, bus departure time, departure intervals and the like; acquiring the running time from the starting station to the terminal station in the similar time period of the previous day and the similar time period of the same characteristic day of the previous week; and acquiring weather characteristics such as weather types, humidity, wind speeds and temperatures when the vehicle runs, and constructing a travel time characteristic data set.
And step S3, based on the correlation coefficient and variance analysis, carrying out correlation analysis on the characteristic factors in the characteristic data set and the travel time, eliminating the characteristic factors which are irrelevant and poor in correlation, matching the characteristic factors with the travel time basic data set, and obtaining a travel time prediction data set.
1. And (3) carrying out correlation analysis on the continuous characteristic (the departure time of the vehicle, the departure interval, the travel time of the similar time period on the last day of the operation shift, the travel time of the same characteristic day of the last week, humidity, wind speed and temperature) and the travel time of the bus by using the Pearson coefficient correlation coefficient.
2. Analysis of variance was performed on the classification characteristics (bus number, driver number, weather type) and travel time of the bus using analysis of variance.
3. And obtaining a travel time prediction data set according to the correlation coefficient and the variance analysis result.
And S4, constructing a double-attention-machine and bidirectional double-layer LSTM neural network bus travel time prediction model based on the attention machine and the LSTM, inputting the travel time prediction data set into the model, predicting the travel time of the bus, and outputting the predicted travel time of the bus.
1. A bus travel time prediction model of a double-attention mechanism and a bidirectional double-layer LSTM is built based on a Kares deep learning framework.
2. Dividing a travel time prediction data set, and using B2 road bus operation data of a weekday late peak from 1/10/2020 to 4/12/2020 for training; b2 bus operation data of weekday peak late at 12 months and 7 days to 12 months and 11 days are used for verification.
3. And inputting the training data in the travel time prediction data set into a bus travel time prediction model of a double-attention machine system and a bidirectional double-layer LSTM for training, and taking the root mean square error as a measurement index.
4. And predicting the travel time of the B2 bus with the late peak of the working day from 7 days in 12 months to 11 days in 12 months, and outputting the predicted travel time.
And (4) following the specific implementation steps, obtaining a comparison graph of the predicted value and the real value of the bus running time of the working day late peak from 7 days in 12 months to 11 days in 12 months. The average absolute error of the predicted time of the bus travel time is 6.23 percent based on the prediction of the invention.
TABLE 1
As shown in table 1, table 1 compares the late peak prediction results of the dual-attention bi-directional dual-layer LSTM model on each working day;
the expressive force of the model is different from Monday to Friday, the prediction precision is higher from Monday to Thursday, the average error is lower than the total average error, the prediction precision is the best on Monday, and the average absolute error is about 3 minutes. However, the prediction precision in friday is obviously reduced, the average absolute percentage error is only 8.88 percent, and the average absolute error is about 7 minutes. The B2 bus line has two main characteristics as can be found by analyzing the B2 bus line. The first characteristic is that the B2 line passes through more universities, which total 7 cents, and respectively: guangzhou industry university, Guangzhou engineering and technology college, Guangzhou sports college, south China university, river-south university, Guangdong post and telecommunications technology college and Guangdong technology college; secondly, the department stores and shopping centers with more routes of B2 are respectively as follows, 16 places in total: shiguangli, Libo square, New big New department, Guangbai department, VT101 dimension Duoli square, Tianhecheng shopping square, Zhengjia square, Wanling Hui, Taigu Hui, modern department, Tian entertainment square, Wal Ma, Jun Tang shopping square, Tianhecheng department, gathering time constellation square, Dongfu garden shopping center. Therefore, during the morning and evening peak hours, more college students and citizens go to various department stores and shopping centers for playing, and the traffic is obviously increased compared with that from monday to thursday. Under the condition, on one hand, the number of passengers getting on and off the bus is obviously increased, the stop time of the bus at each stop is changed, and the complexity of the time-space law of operation is increased; on the other hand, the traffic volume of private cars and network appointments increases, which leads to more complex road conditions of the operation route where the buses are located. Therefore, when the model predicts the travel time of the buses with the peak at friday night, the prediction difficulty is high, the time-space law of the bus operation is difficult to capture in a complex traffic environment, the precision is reduced to some extent, the average error is still lower than 9%, and the bus scheduling requirement of a bus company is met overall.
In this embodiment, the dual-attention bidirectional double-layer LSTM model is compared with other neural network models, and table 2 shows the comparison of the dual-attention bidirectional double-layer LSTM model with other neural network model experiments.
TABLE 2
Specifically, in order to better analyze the improvement effect of the feature importance extraction module, the time feature module and the travel time prediction module on the travel time prediction precision, the prediction precision of the model is compared from two aspects. On the one hand, this model is compared with other neural network models: compared with a single-attention bidirectional double-layer LSTM model, the relative precision is improved by 7.98%; compared with a non-attention bidirectional double-layer LSTM model, the relative precision is improved by 16.3%; compared with the double-layer LSTM, the relative precision is improved by 19.3%. On the other hand, compared with the travel time prediction model of the traditional method, the method comprises the following steps: compared with a multilayer perceptron, the relative precision is improved by 29.0%; compared with the support vector machine model, the relative precision is improved by 31.0%. According to the comparison result, the feature importance extraction module and the time feature module have great improvement on the accuracy of the prediction of the bus travel time, can mine deep time features in the bus running process, and can more pertinently integrate different features into model prediction to obtain higher prediction accuracy.
The above description is only a preferred embodiment of the present invention, and all equivalent changes and modifications made in accordance with the claims of the present invention should be covered by the present invention.
Claims (6)
1. A bus travel time prediction method based on a double-attention mechanism and a bidirectional double-layer LSTM is characterized by comprising the following steps:
s1, acquiring bus attribute data and station entering and exiting data, preprocessing the data and constructing a travel time basic data set;
s2, acquiring basic characteristics, running time and weather characteristics of the bus during running, and constructing a travel time characteristic data set;
step S3, based on correlation coefficient and variance analysis, carrying out correlation analysis on the characteristic factors in the travel time characteristic data set and the travel time, eliminating the characteristic factors which are irrelevant and have poor correlation, and matching with the travel time basic data set to obtain a travel time prediction data set;
and S4, constructing a double-attention machine mechanism and a bidirectional double-layer LSTM neural network bus travel time prediction model, inputting the travel time prediction data set into the model, predicting the travel time of the bus, and outputting the predicted travel time of the bus.
2. The method for predicting the bus journey time based on the double-attention mechanism and the bidirectional double-layer LSTM according to claim 1, wherein the step S1 is specifically as follows:
s11, acquiring bus attribute data and station entering and exiting data, directly eliminating repeated records and abnormal data and complementing missing data;
and step S12, matching the attribute data of the bus with the station entering and exiting data according to the serial number of the bus operation shifts, acquiring the bus operation time of each operation shift, and constructing a travel time basic data set.
3. The method for predicting the bus journey time based on the double-attention mechanism and the bidirectional double-layer LSTM according to claim 1, wherein the step S2 is specifically as follows: acquiring a bus number, a driver number, bus departure time and a bus departure interval of a bus; acquiring the running time from the starting station to the terminal station in the similar time period of the previous day and the similar time period of the same characteristic day of the previous week; and acquiring weather characteristics including weather type, humidity, wind speed and temperature when the vehicle runs, and constructing a travel time characteristic data set.
4. The method for predicting the bus journey time based on the double-attention mechanism and the bidirectional double-layer LSTM according to claim 1, wherein the step S3 is specifically as follows:
step S31: performing correlation analysis between the continuous characteristic factors and the bus running time by using the correlation coefficient of the Pearson correlation coefficient to obtain the correlation between variables;
step S32: classifying and coding the weather type, the driver number and the bus number discrete characteristic factor, converting the characteristic factors into classification characteristic factors, and analyzing the correlation between the classification characteristic factors and the bus running time by using variance test;
step S33: and (4) discarding the characteristic factors which are irrelevant and have poor correlation, and matching the retained characteristics with the travel time basic data set to obtain a travel time prediction data set.
5. The method for predicting the bus journey time based on the double-attention mechanism and the bidirectional double-layer LSTM according to claim 1, wherein the step S4 is specifically as follows:
step S41: inputting the travel time prediction data set into a feature importance extraction module, and distributing attention of different degrees to the features according to the importance difference of the features to obtain a feature importance matrix;
step S42: splicing the characteristic importance matrix with the input characteristics of the model to obtain a bus running state matrix;
step S43: inputting the bus running state matrix into a time characteristic extraction module, extracting time characteristics in the bus running process by using a bidirectional LSTM, and generating a running time characteristic matrix;
step S44: inputting the obtained running time characteristic matrix into a travel time prediction module, fusing Attention into the time step of the LSTM, and constructing an Attention _ LSTM layer to realize extraction of influence characteristics of buses with different time distances; in the full-connection layer, the root mean square error is used as a loss function, the prediction result is compared with the actual result, and the model is continuously trained; the prediction of the bus travel time is realized.
6. A bus travel time prediction system based on a double-attention mechanism and a bidirectional double-layer LSTM is characterized by comprising
The feature importance extraction module is used for acquiring importance weights of all features by using an attention mechanism to generate a feature importance matrix, distinguishing the influence of different features on the travel time of the bus and improving the prediction efficiency and the prediction precision of the model;
the time characteristic extraction module is used for capturing the running characteristics of the buses running before and after by utilizing the stronger time characteristic capture capability of the bidirectional LSTM, and acquiring the time characteristics of the buses in the running process;
and the travel time prediction module introduces Attention into the LSTM, constructs an Attention _ LSTM layer, gives different Attention to different time steps in the LSTM and expresses the influence of vehicles at different time distances on the travel time of the bus.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210007787.9A CN114328675A (en) | 2022-01-06 | 2022-01-06 | Bus travel time prediction method based on double-attention machine system and bidirectional double-layer LSTM |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210007787.9A CN114328675A (en) | 2022-01-06 | 2022-01-06 | Bus travel time prediction method based on double-attention machine system and bidirectional double-layer LSTM |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114328675A true CN114328675A (en) | 2022-04-12 |
Family
ID=81024754
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210007787.9A Pending CN114328675A (en) | 2022-01-06 | 2022-01-06 | Bus travel time prediction method based on double-attention machine system and bidirectional double-layer LSTM |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114328675A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114881303A (en) * | 2022-04-21 | 2022-08-09 | 海南大学 | Water quality long-term prediction method based on double-attention machine system |
| CN115565376A (en) * | 2022-09-30 | 2023-01-03 | 福州大学 | Vehicle travel time prediction method and system fusing graph2vec and double-layer LSTM |
| CN116596126A (en) * | 2023-04-27 | 2023-08-15 | 苏州大学 | Bus string prediction method and system |
| CN117192063A (en) * | 2023-11-06 | 2023-12-08 | 山东大学 | Water quality prediction method and system based on coupled Kalman filtering data assimilation |
-
2022
- 2022-01-06 CN CN202210007787.9A patent/CN114328675A/en active Pending
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114881303A (en) * | 2022-04-21 | 2022-08-09 | 海南大学 | Water quality long-term prediction method based on double-attention machine system |
| CN114881303B (en) * | 2022-04-21 | 2024-06-04 | 海南大学 | Water quality long-term prediction method based on double-attention mechanism |
| CN115565376A (en) * | 2022-09-30 | 2023-01-03 | 福州大学 | Vehicle travel time prediction method and system fusing graph2vec and double-layer LSTM |
| CN115565376B (en) * | 2022-09-30 | 2024-05-03 | 福州大学 | Vehicle journey time prediction method and system integrating graph2vec and double-layer LSTM |
| CN116596126A (en) * | 2023-04-27 | 2023-08-15 | 苏州大学 | Bus string prediction method and system |
| CN117192063A (en) * | 2023-11-06 | 2023-12-08 | 山东大学 | Water quality prediction method and system based on coupled Kalman filtering data assimilation |
| CN117192063B (en) * | 2023-11-06 | 2024-03-15 | 山东大学 | Water quality prediction method and system based on coupled Kalman filtering data assimilation |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114328675A (en) | Bus travel time prediction method based on double-attention machine system and bidirectional double-layer LSTM | |
| CN109034566B (en) | A kind of intelligent dispatching method and device based on passenger flow above and below bus station | |
| CN102799897B (en) | Computer recognition method of GPS (Global Positioning System) positioning-based transportation mode combined travelling | |
| CN104123833B (en) | A kind of planning method and device of condition of road surface | |
| CN104318324B (en) | Shuttle Bus website and route planning method based on taxi GPS records | |
| CN105513337B (en) | The Forecasting Methodology and device of a kind of volume of the flow of passengers | |
| CN113177657B (en) | Rail transit passenger flow prediction method and device | |
| CN110400462B (en) | Track traffic passenger flow monitoring and early warning method and system based on fuzzy theory | |
| Gong et al. | Hybrid dynamic prediction model of bus arrival time based on weighted of historical and real-time GPS data | |
| CN102324128A (en) | Method for predicting OD (Origin-Destination) passenger flow among bus stations on basis of IC (Integrated Circuit)-card record and device | |
| CN110299011A (en) | A kind of traffic flow forecasting method of the highway arbitrary cross-section based on charge data | |
| CN110298516B (en) | Method and device for splitting overlong bus line based on passenger flow OD data, mobile terminal equipment and server | |
| Wilson et al. | The potential impact of automated data collection systems on urban public transport planning. | |
| CN111932925A (en) | Method, device and system for determining travel passenger flow of public transport station | |
| CN110853156B (en) | Passenger OD identification method integrating bus GPS track and IC card data | |
| CN111063204B (en) | Expressway vehicle speed prediction model training method based on toll station flow | |
| WO2021115320A1 (en) | Traffic evaluation method and system | |
| CN111414719B (en) | Method and device for extracting peripheral features of subway station and estimating traffic demand | |
| CN115063978B (en) | Bus arrival time prediction method based on digital twins | |
| CN108062857A (en) | For the Forecasting Methodology of cab-getter's trip purpose | |
| CN115018148A (en) | Urban rail transit network passenger flow distribution prediction method and system based on digital twin model | |
| CN113380043B (en) | Bus arrival time prediction method based on deep neural network calculation | |
| CN112863182A (en) | Cross-modal data prediction method based on transfer learning | |
| CN112085048A (en) | Vehicle frequent track identification method and device based on LSTM self-encoding | |
| CN116523093A (en) | Grid demand sensing system and method of energy system based on random source load prediction |
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 |