Search

CN-120013118-B - Bus commuter passenger travel demand analysis method based on space-time characteristics

CN120013118BCN 120013118 BCN120013118 BCN 120013118BCN-120013118-B

Abstract

The invention discloses a bus commuter passenger travel demand analysis method based on space-time characteristics, which comprises the steps of obtaining historical travel data of bus passengers and constructing a personal travel behavior data set, screening a gate commuter passenger group based on travel frequency, analyzing travel space-time rules according to the travel time of the passengers, entropy values of travel stations and travel chain similarity, extracting travel chain dimension characteristics, time dimension characteristics and space dimension characteristics related to a prediction target according to the travel space-time rules, constructing a characteristic matrix based on the extracted characteristics, training the characteristics in the characteristic matrix by using a distributed gradient enhancement model, and predicting and outputting travel demands of the commuter passengers based on the trained travel demand prediction model. The invention realizes the efficient and accurate prediction of the travel demands of the commuter passengers by integrating the multidimensional travel characteristics of the commuter passengers and the machine learning algorithm.

Inventors

  • ZOU LIANG
  • LI SHAO
  • QIU JIAQI

Assignees

  • 深圳大学

Dates

Publication Date
20260505
Application Date
20241220

Claims (7)

  1. 1. A bus commuter passenger travel demand analysis method based on space-time characteristics is characterized by comprising the following steps: Acquiring historical trip data of bus passengers, and constructing a personal trip behavior data set; screening and selecting a service passenger group based on the travel frequency, and analyzing travel space-time rules according to the travel time of the passengers, the entropy value of travel stations and travel chain similarity; Extracting travel chain dimension characteristics, time dimension characteristics and space dimension characteristics related to a prediction target according to the travel space-time law; constructing a feature matrix based on the extracted features, training the features in the feature matrix by using a distributed gradient enhancement model, and predicting and outputting travel demands of commuter passengers based on a trained travel demand prediction model; the step of extracting travel chain dimension features, time dimension features and space dimension features related to the prediction target according to the travel space-time law comprises the following steps: Extracting OD 1 data with the same travel sequence on the same workday, travel time T 1 with the same travel sequence on the same workday, highest-frequency travel time T 2 corresponding to the OD 1 data and highest-frequency travel time OD 2 corresponding to the travel time T 1 data to obtain the travel chain dimension characteristics; extracting the travel time of commuter passengers in the last preset travel time to obtain the time dimension characteristics; Extracting travel OD data corresponding to the last preset travel of the commuter, extracting highest frequency OD data corresponding to a history starting point O with the shortest distance between each travel and a previous travel end point D, and obtaining the space dimension characteristics; the feature matrix is constructed based on the extracted features, the features in the feature matrix are trained by using a distributed gradient enhancement model, and travel demands of commuter passengers are predicted and output based on a trained travel demand prediction model, and the method comprises the following steps: Constructing the feature matrix based on the extracted features, and performing independent heat coding on travel chain dimension features, time dimension features and space dimension features in the feature matrix; Training travel chain dimension features, time dimension features and space dimension features in the feature matrix based on the distributed gradient enhancement model to obtain a trained travel demand prediction model; predicting and outputting travel demands of commuter passengers based on the trained travel demand prediction model; the training of the travel chain dimension feature, the time dimension feature and the space dimension feature in the feature matrix based on the distributed gradient enhancement model comprises the following steps: training the extracted space-time characteristics based on the distributed gradient enhancement model to construct a travel demand prediction model; and verifying the travel demand prediction model by adopting a K-fold cross verification method, and optimizing model parameters according to verification results to obtain the trained travel demand prediction model.
  2. 2. The method for analyzing travel demands of bus commuter passengers based on space-time characteristics according to claim 1, wherein the steps of obtaining historical travel data of bus passengers and constructing a personal travel behavior data set comprise the following steps: Acquiring information of a boarding station, a alighting station, boarding time and alighting time of bus passengers, and acquiring the historical trip data; and constructing the personal trip behavior data set according to the historical trip data.
  3. 3. The method for analyzing travel demand of bus commute passengers based on space-time characteristics according to claim 1, wherein the screening of the population of commute passengers based on travel frequency comprises: Analyzing daily trip frequency, zhou Chuhang frequency and repeated trip chain trip frequency of corresponding passengers based on the personal trip behavior data set; And screening the commuter passengers from the historical travel data of the bus passengers according to the daily travel frequency, the Zhou Chuhang frequency and the repeated travel chain travel frequency, and determining the commuter passenger group.
  4. 4. The method for analyzing travel demands of bus commute passengers based on space-time characteristics according to claim 1, wherein the analyzing travel space-time rule according to the travel time of passengers, the entropy value of travel stations and travel chain similarity comprises the following steps: And respectively calculating the travel time, the entropy value of the travel station and the travel chain similarity of the same workday corresponding to the commuter passengers and the non-commuter passengers, and comparing to obtain the travel mode space-time rule of the commuter passengers.
  5. 5. A space-time feature-based bus commute passenger travel demand analysis system for implementing the space-time feature-based bus commute passenger travel demand analysis method as set forth in any one of claims 1 to 4, comprising: The data set acquisition module is used for acquiring historical trip data of bus passengers and constructing a personal trip behavior data set; The travel time-space rule analysis module is used for screening the gate duty passenger groups based on travel frequency and analyzing travel time-space rules according to the travel time of the passengers, the entropy value of travel stations and travel chain similarity; The multidimensional feature extraction module is used for extracting travel chain dimension features, time dimension features and space dimension features related to the prediction targets according to the travel space-time law; The prediction and output module is used for constructing a feature matrix based on the extracted features, training the features in the feature matrix by using the distributed gradient enhancement model, and predicting and outputting the travel demands of commuter passengers based on the trained travel demand prediction model.
  6. 6. A terminal comprising a processor and a memory, wherein the memory stores a space-time feature based bus commuter passenger travel demand analysis program, which when executed by the processor is operative to implement the space-time feature based bus commuter passenger travel demand analysis method of any one of claims 1-4.
  7. 7. A computer readable storage medium, wherein the computer readable storage medium stores a space-time feature based bus commuter passenger travel demand analysis program which, when executed by a processor, is operable to implement the space-time feature based bus commuter passenger travel demand analysis method of any one of claims 1-4.

Description

Bus commuter passenger travel demand analysis method based on space-time characteristics Technical Field The invention relates to the technical field of public transportation big data processing, in particular to a method for analyzing travel demands of bus commuter passengers based on space-time characteristics. Background In recent years, urban buses face a serious challenge of continuous decrease in passenger traffic. According to the data, the passenger traffic of the conventional public transportation system is reduced year by year from 2012 to 2022, and meanwhile, the daily passenger traffic of a single bus is also reduced remarkably. This results in a huge contrast between the high capacity investment and the low passenger flow efficiency of urban public traffic. To address this challenge, public transportation companies have begun taking various measures to optimize operating efficiency and quality of service. Methods of optimizing line layout, adding custom bus service, adjusting the number and distribution of vehicles, etc. have been widely used in large cities. For example, city a optimizes 138 bus routes at the end of 2022, adds 60 stations, and realizes 200 meters connection between all subway stations in the urban center and the bus routes. Although these measures alleviate operating pressure to some extent, more accurate predictions of passenger demand are needed to ensure efficient allocation and continued improvement of bus resources. At present, demand prediction in the public transportation industry depends on a traditional historical data analysis method, and the time-space characteristics and travel chain rules of the travel behaviors of passengers are ignored. Especially commuter passenger demands are often concentrated in the morning and evening peak hours, and it is difficult for conventional methods to accurately predict these fluctuating demand changes. With increasing complexity of urban space structures and diversification of travel modes, the conventional demand prediction method cannot completely meet the capacity scheduling demands of a bus system in a peak period. This will result in a mismatch in capacity and demand and will not provide efficient and convenient travel services for the passengers. Accordingly, there is a need in the art for improvement. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a bus commuter passenger travel demand analysis method based on space-time characteristics, so as to solve the problem that the traditional historical data analysis method is difficult to accurately predict the fluctuation drivability scheduling demand. The technical scheme adopted for solving the technical problems is as follows: In a first aspect, the present invention provides a method for analyzing travel demands of passengers on bus commuter based on space-time characteristics, comprising: Acquiring historical trip data of bus passengers, and constructing a personal trip behavior data set; screening and selecting a service passenger group based on the travel frequency, and analyzing travel space-time rules according to the travel time of the passengers, the entropy value of travel stations and travel chain similarity; Extracting travel chain dimension characteristics, time dimension characteristics and space dimension characteristics related to a prediction target according to the travel space-time law; And constructing a feature matrix based on the extracted features, training the features in the feature matrix by using a distributed gradient enhancement model, and predicting and outputting the travel demands of commuter passengers based on the trained travel demand prediction model. In one implementation manner, the obtaining the historical trip data of the bus passengers and constructing the personal trip behavior data set includes: Acquiring information of a boarding station, a alighting station, boarding time and alighting time of bus passengers, and acquiring the historical trip data; and constructing the personal trip behavior data set according to the historical trip data. In one implementation, the screening the population of duty passengers based on travel frequency includes: Analyzing daily trip frequency, zhou Chuhang frequency and repeated trip chain trip frequency of corresponding passengers based on the personal trip behavior data set; And screening the commuter passengers from the historical travel data of the bus passengers according to the daily travel frequency, the Zhou Chuhang frequency and the repeated travel chain travel frequency, and determining the commuter passenger group. In one implementation manner, the analyzing the travel time-space rule according to the travel time of the passenger, the entropy value of the travel station and the travel chain similarity includes: And respectively calculating the travel time, the entropy value of the travel station and the travel chain similarity of the same workday corre