Bus commuting passenger travel demand analysis method based on spatial-temporal characteristics

Abstract

The invention discloses a bus commuting passenger travel demand analysis method based on spatial-temporal characteristics, and the method comprises the steps: obtaining the historical travel data of bus passengers, and constructing a personal travel behavior data set; screening commuting passenger groups based on the travel frequency, and analyzing a travel time-space law according to the travel time of the passengers, the entropy value of the travel station and the travel chain similarity; extracting travel chain dimension features, time dimension features and space dimension features related to a prediction target according to the travel time-space 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 the commuting passengers based on the trained travel demand prediction model. According to the method, the multi-dimensional travel characteristics of the commuting passengers and the machine learning algorithm are fused, so that the travel demands of the commuting passengers are efficiently and accurately predicted.

Description

Technical Field

[0001] The present invention relates to the technical field of public transportation big data processing, and in particular to a method for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics. Background Art

[0002] In recent years, urban public transportation has faced the severe challenge of a continuous decline in passenger volume. According to data, from 2012 to 2022, the passenger volume of conventional public transportation systems has declined year by year, and the daily passenger volume of a single vehicle has also declined significantly. As a result, urban public transportation faces a huge contrast between high capacity investment and low passenger flow benefits.

[0003] In order to meet this challenge, bus companies have begun to take various measures to optimize operational efficiency and service quality. Methods such as optimizing route layout, increasing customized bus services, and adjusting the number and distribution of vehicles have been widely used in many large cities. For example, City A optimized 138 bus routes, added 60 new stops, and achieved a 200-meter connection between all subway stations and bus routes in the city center by the end of 2022. Although these measures have alleviated operational pressure to a certain extent, more accurate passenger demand forecasts are still needed to ensure the effective allocation and continuous improvement of bus resources.

[0004] At present, demand forecasting in the public transportation industry mostly relies on traditional historical data analysis methods, ignoring the spatiotemporal characteristics of passenger travel behavior and the laws of travel chains. In particular, the demand of commuting passengers is usually concentrated in the morning and evening peak hours, and traditional methods are difficult to accurately predict the changes in these fluctuating demands. With the increasing complexity of urban spatial structure and the diversification of travel modes, traditional demand forecasting methods cannot fully meet the capacity scheduling needs of public transportation systems during peak hours. This will lead to a mismatch between capacity and demand, and it will be impossible to provide passengers with efficient and convenient travel services.

[0005] Therefore, the prior art needs to be improved. Summary of the invention

[0006] The technical problem to be solved by the present invention is that, in view of the defects of the existing technology, the present invention provides a method for analyzing the travel demand of public transportation commuters based on spatiotemporal characteristics, so as to solve the problem that traditional historical data analysis methods are difficult to accurately predict the fluctuating capacity scheduling demand.

[0007] The technical solution adopted by the present invention to solve the technical problem is as follows: In a first aspect, the present invention provides a method for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics, comprising: Obtain historical travel data of public transport passengers and construct a personal travel behavior dataset; The commuter passenger groups are screened based on the travel frequency, and the travel time and space patterns are analyzed according to the entropy value of the passengers' travel time and travel stations, and the similarity of the travel chain; Extracting travel chain dimension features, time dimension features and space dimension features related to the prediction target according to the travel time and space law; A feature matrix is ​​constructed based on the extracted features, the features in the feature matrix are trained using a distributed gradient enhancement model, and the travel demand of commuting passengers is predicted and output based on the trained travel demand prediction model.

[0008] In one implementation, the acquisition of historical travel data of public transport passengers and the construction of a personal travel behavior dataset include: Obtain the boarding station, alighting station, boarding time and alighting time information of the bus passengers to obtain the historical travel data; The personal travel behavior dataset is constructed according to the historical travel data.

[0009] In one implementation, screening commuter passenger groups based on travel frequency includes: Analyze the daily travel frequency, weekly travel frequency, and repeated travel chain travel frequency of the corresponding passenger based on the personal travel behavior data set; According to the daily travel frequency, the weekly travel frequency and the repeated travel chain travel frequency, commuter passengers are screened from the historical travel data of the bus passengers to determine the commuter passenger group.

[0010] In one implementation, analyzing the travel time and space regularity according to the entropy value of the passenger's travel time and travel stops and the similarity of the travel chain includes: The travel time, entropy values ​​of travel sites and similarity of travel chains on the same weekday corresponding to commuting passengers and non-commuting passengers are calculated respectively, and the spatiotemporal regularity of the travel patterns of the commuting passengers is obtained by comparison.

[0011] In one implementation, the extracting of travel chain dimension features, time dimension features, and space dimension features related to the prediction target according to the travel time-space law includes: Extract the OD1 data of the same travel order on the same working day, the travel time T1 of the same travel order on the same working day, the most frequent travel time T2 corresponding to the OD1 data, and the most frequent travel time OD2 data corresponding to the travel time T1, to obtain the travel chain dimensional features; Extracting the travel time of the commuter passengers in the most recent preset trip to obtain the time dimension feature; The commuter passenger's travel OD data corresponding to the most recent preset trip is extracted, and the highest frequency OD data corresponding to the historical starting point O that is closest to the previous trip end point D of each trip is extracted to obtain the spatial dimension feature.

[0012] In one implementation, constructing a feature matrix based on the extracted features, training the features in the feature matrix using a distributed gradient enhancement model, and predicting and outputting the travel demand of commuting passengers based on the trained travel demand prediction model include: The feature matrix is ​​constructed based on the extracted features, and the travel chain dimension features, time dimension features, and space dimension features in the feature matrix are uniquely encoded; Based on the distributed gradient enhancement model, the travel chain dimension features, the time dimension features and the space dimension features in the feature matrix are trained to obtain the trained travel demand prediction model; The travel demand of commuting passengers is predicted and output based on the trained travel demand prediction model.

[0013] In one implementation, the training of the travel chain dimension features, the time dimension features, and the space dimension features in the feature matrix based on the distributed gradient enhancement model includes: Based on the distributed gradient enhancement model, the extracted spatiotemporal features are trained to build a travel demand prediction model; The travel demand forecasting model is verified by using a K-fold cross-validation method, and the model parameters are optimized according to the verification results to obtain the trained travel demand forecasting model.

[0014] In a second aspect, the present invention provides a public transportation commuter travel demand analysis system based on spatiotemporal characteristics, comprising: The dataset acquisition module is used to obtain the historical travel data of bus passengers and construct a personal travel behavior dataset; The travel time-space law analysis module is used to screen commuter passenger groups based on travel frequency, and analyze the travel time-space law based on the passengers' travel time, entropy value of travel stations, and travel chain similarity; A multi-dimensional feature extraction module, used to extract travel chain dimensional features, time dimensional features and space dimensional features related to the prediction target according to the travel time and space law; The prediction and output module is used to construct a feature matrix based on the extracted features, train the features in the feature matrix using a distributed gradient enhancement model, and predict and output the travel demand of commuting passengers based on the trained travel demand prediction model.

[0015] In a third aspect, the present invention provides a terminal comprising: a processor and a memory, wherein the memory stores a program for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics, and when the program for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics is executed by the processor, it is used to implement the operation of the method for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics as described in the first aspect.

[0016] In a fourth aspect, the present invention further provides a medium, which is a computer-readable storage medium, and which stores a public transportation commuter travel demand analysis program based on spatiotemporal characteristics, and when the public transportation commuter travel demand analysis program based on spatiotemporal characteristics is executed by a processor, it is used to implement the operation of the public transportation commuter travel demand analysis method based on spatiotemporal characteristics as described in the first aspect.

[0017] The present invention adopts the above technical solution to achieve the following effects: By analyzing the historical travel data of public transportation commuters, the present invention can determine the spatiotemporal patterns of commuting passengers, and by extracting the features of time, space and travel chain dimensions, the XGBoost algorithm is used to model the multidimensional features, thereby constructing a machine learning prediction model, which can effectively improve the accuracy of demand forecasting and thus accurately predict the travel needs of commuting passengers. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying creative work.

[0019] Figure 1 It is a flow chart of the method for analyzing travel demand of public transportation commuter passengers based on spatiotemporal characteristics in the present invention.

[0020] Figure 2 It is a similarity heat map of the travel chain of a commuting passenger during the working day in the present invention.

[0021] Figure 3 It is a schematic diagram of the high-frequency travel OD distribution of a commuter passenger in the present invention.

[0022] Figure 4 It is a schematic diagram of the system structure of the method for analyzing the travel demand of public transportation commuters based on spatiotemporal characteristics in the present invention.

[0023] Figure 5 It is a functional principle diagram of a terminal in one implementation of the present invention.

[0024] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings in conjunction with the embodiments. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solution and advantages of the present invention clearer and more specific, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0026] Exemplary Methods The current passenger flow forecasting of urban public transportation systems mainly relies on traditional traffic demand models, which usually consider time and space characteristics, but these methods fail to make full use of the historical travel data of individual passengers for efficient and accurate forecasting. Especially for the demand forecasting of commuter passengers, due to the high regularity and spatiotemporal characteristics of their travel behavior, traditional methods often ignore these characteristics, resulting in poor prediction results. Therefore, it is necessary to adopt a new method for predicting the travel demand of public transportation commuter passengers based on time, space and travel chain characteristics.

[0027] In response to the above technical problems, the embodiment of the present invention provides a method for predicting the travel demand of public transportation commuters based on time, space and travel chain characteristics. The method deeply analyzes the bus card swiping data, extracts the key factors affecting the travel demand of commuters from the multi-dimensional characteristics of time regularity, spatial matching and historical travel chain, and realizes accurate prediction of the future travel behavior of individual commuters based on machine learning algorithms.

[0028] The embodiment of the present invention first identifies the commuter passenger group and reveals the significant regularity of their travel behavior in time and space. In the time dimension, the travel time of commuter passengers has a higher regularity; in the space dimension, the distribution of the starting and ending stations of commuter passengers is more concentrated and stable. Based on these characteristics, the embodiment of the present invention comprehensively considers the order characteristics of the historical travel chain and its relevance to future travel, extracts the characteristics of the three dimensions of time, space, and travel chain, and establishes a data-driven demand forecasting framework. By introducing the XGBoost model (distributed gradient boosting model), combined with feature engineering optimization and model validation evaluation, the embodiment of the present invention effectively improves the accuracy and reliability of the prediction, realizes the efficient prediction of the next travel time and starting and ending points of individual commuter passengers, and significantly improves the prediction accuracy.

[0029] like Figure 1 As shown, an embodiment of the present invention provides a method for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics, comprising the following steps: Step S100, obtaining historical travel data of public transport passengers and constructing a personal travel behavior dataset.

[0030] In this embodiment, it is necessary to first obtain historical travel data including information such as the departure place, destination, and travel time to construct a personal travel behavior data set; the public transportation system is an important embodiment of urban operating efficiency and sustainable development, especially in urban commuting, the matching degree between the operating efficiency of the public transportation system and passenger demand directly affects the optimization of urban traffic management and the travel experience of residents. Therefore, accurately obtaining the travel behavior data of passengers and constructing a reasonable personal travel behavior data set are important for the subsequent further analysis of passengers.

[0031] In this embodiment, the smart card swipe data is combined with the bus AVL trajectory (automatic vehicle location trajectory) data to obtain the passenger's travel origin, destination, travel time and other information. Through the collection and processing of historical travel data, the boarding and alighting stations (OD), boarding time, alighting time, bus routes and the like are obtained.

[0032] Specifically, in an implementation of this embodiment, step S100 includes the following steps: Step S101, obtaining the boarding station, alighting station, boarding time and alighting time information of the bus passengers to obtain the historical travel data; Step S102: construct the personal travel behavior dataset based on the historical travel data.

[0033] In this embodiment, data from the bus card swiping system is obtained, including information such as each passenger's boarding and alighting time, bus number, starting station and destination station, and through data cleaning and processing, a passenger data set that meets the demand analysis is obtained.

[0034] As an example, in an actual application scenario, first, the historical card swiping data of the bus system in City A is obtained, the card swiping time and the corresponding boarding station data of each passenger are extracted, and the alighting station and the corresponding alighting time information of the travel record are obtained by matching the bus AVL trajectory data; then, based on the passenger's historical travel records, the OD data pairs (departure and destination) and their corresponding time information of each trip are analyzed to construct a personal travel behavior dataset; based on the personal travel behavior dataset of each passenger, the structural characteristics of the travel chain (continuous travel chain and time series) are extracted for subsequent analysis; finally, the extracted OD data pairs are matched with the geographic information data of the map platform (for example, a certain German map data), so as to calibrate the coordinates and location labels of the departure and destination, and obtain an accurate personal travel behavior dataset for each passenger.

[0035] like Figure 1 As shown, an embodiment of the present invention provides a method for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics, comprising the following steps: Step S200, screening commuter passenger groups based on travel frequency, and analyzing travel time and space patterns according to the passengers' travel time, entropy values ​​of travel sites, and travel chain similarity.

[0036] In this embodiment, commuter passenger groups are screened based on travel frequency. Commuting travel usually has the characteristics of high frequency and strong regularity. By exploring the patterns of travel behavior in time and space distribution, it can provide key input variables for subsequent demand forecasting modeling.

[0037] Specifically, in an implementation of this embodiment, step S200 includes the following steps: Step S201, analyzing the daily travel frequency, weekly travel frequency and repeated travel chain travel frequency of the corresponding passenger based on the personal travel behavior data set; Step S202, screening commuter passengers from the historical travel data of the public transport passengers according to the daily travel frequency, the weekly travel frequency and the repeated travel chain travel frequency, and determining the commuter passenger group; Step S203, respectively calculating the travel time corresponding to commuting passengers and non-commuting passengers, the entropy value of the travel sites, and the similarity of the travel chains on the same weekday, and comparing to obtain the spatiotemporal regularity of the travel patterns of the commuting passengers.

[0038] In this embodiment, after completing the extraction of the personal travel behavior data set, this embodiment accurately identifies the commuting passenger group by setting the screening rules for commuting behavior. Specifically, according to the travel chain structure within a week, if the same travel chain (including OD and time combination) of a passenger is repeated twice or more on different working days, it can be determined that it has significant repetitive behavior. In addition, passengers with an average daily frequency of two or more trips are screened to ensure that they meet the high-frequency characteristics of commuting behavior. At the same time, combined with the condition that the total number of trips in a week reaches three or more, the stability and regularity of commuting passengers are further limited.

[0039] In this embodiment, based on the identified commuter passenger groups, the screened commuter passenger data is used to further analyze their spatiotemporal regularity. From the time dimension, the time regularity of commuter travel chains on different weekdays and the regularity of passenger travel time periods are analyzed.

[0040] Specifically, when analyzing the temporal regularity of commuting trip chains on different weekdays and the regularity of passenger travel time periods, we first conduct a time series analysis of the trip chains on each weekday of the personal travel behavior dataset to study the similarities of the trip chains on different weekdays. The specific steps are as follows: 1. Using the Jaccard similarity index, we calculated the overlap ratio of passenger travel stops on different weekdays and measured the similarity of spatial distribution; In this embodiment, the Jaccard similarity of the two working day trip chains is expressed as follows:

[0041] Represents the OD set of chain A; B represents the OD set of trip chain B.

[0042] 2. Use the edit distance algorithm to match and compare the sequence of travel chain sites, analyze the consistency of the order of each travel chain, and eliminate the influence of chain length differences through standardization; Figure 2 As shown, Figure 2 is a heat map of the similarity of the travel chains of a commuter in City A during the working day. Figure 2 The similarity thermal value of the travel chains between working days can be used to analyze the consistency of the order of each travel chain.

[0043] 3. Combining Jaccard similarity and standardized edit distance, a comprehensive similarity model is constructed to characterize the regularity of commuting travel chains on different working days as a whole.

[0044] Then, by analyzing the time periods of the personal travel behavior dataset, the regularity of commuter passengers’ travel time periods is studied. The specific steps are as follows: 1. Extract travel time periods: Extract the boarding time of each travel record from the personal travel behavior dataset and divide it into several time periods according to the specific time, such as hourly division (such as 6:00-7:00, 7:00-8:00, etc.). Multiple travel records of each passenger will be mapped to the corresponding time period.

[0045] 2. Statistical time period frequency: For each passenger, count the travel frequency of all their travel records in different time periods, normalize the frequency, and calculate the travel probability distribution of each time period , which indicates A specific period of time.

[0046] 3. Calculate the entropy value of the travel time period: According to the information entropy formula, calculate the regularity index (entropy value) of the passenger's travel time:

[0047] in: Indicates the passenger's time period The probability of traveling; H represents the entropy value of the passenger's travel time distribution, which measures the uncertainty of the time distribution.

[0048] 4. Determine the regularity of travel time: When the entropy value is low, it means that the passenger's travel time period is relatively concentrated and highly regular; when the entropy value is high, it means that the passenger's travel time distribution is relatively dispersed and the regularity is weak.

[0049] In this embodiment, after analyzing the time regularity of commuting travel chains on different weekdays and the regularity of passenger travel time periods, the origin and destination (OD) of each travel record in the personal travel behavior dataset are analyzed to study the regularity of commuting passengers' travel OD. The specific implementation process is as follows: 1. Extract trip OD data: Extract the departure station (starting point O) and arrival station (end point D) of each trip record from the personal travel behavior dataset and combine them into a trip OD pair (O, D). Multiple trip records of each passenger will be mapped to multiple OD pairs.

[0050] 2. Count the frequency of OD pairs: For each passenger, count the frequency of different OD pairs in all their travel records, normalize the frequencies, and calculate the travel probability distribution of each OD pair. ,in Indicates a specific OD pair.

[0051] 3. Calculate the entropy value of the trip OD: According to the information entropy formula, calculate the regularity index (entropy value) of the passenger in the OD distribution:

[0052] in: Is the passenger's choice of OD The probability of traveling; H represents the entropy value of the passenger's trip OD distribution, which measures the uncertainty of OD selection.

[0053] 4. Determine the regularity of travel OD: When the entropy value is low, it means that the passenger's travel OD is more concentrated and highly regular; when the entropy value is high, it means that the passenger's travel OD selection is more dispersed and the regularity is weak.

[0054] In this embodiment, after analyzing the starting point and end point of each travel record in the personal travel behavior data set, the regularity index results of the time and space dimensions are integrated to obtain the main spatiotemporal regularity of commuting passengers; the specific implementation process is as follows: 1. Comprehensive time and space regularity index: The calculated travel time regularity index (time entropy value on travel time) ) and the calculated travel space regularity index (regularity spatial entropy value on OD distribution ) to comprehensively analyze the regular characteristics of commuter passengers.

[0055] 2. Summarize the time and space patterns of commuting passengers: Time dimension: travel time entropy of commuting passengers It is relatively low, indicating that the distribution of commuting travel time is highly concentrated, usually concentrated in fixed morning and evening peak hours, showing clear time regularity.

[0056] Spatial dimension: OD entropy of commuting passengers It is relatively low, indicating that the spatial distribution of commuting trips is also significantly concentrated, usually corresponding to fixed departure and destination stations, showing a stable spatial regularity.

[0057] 3. Output the main spatiotemporal regularity of commuting passengers: The overall travel behavior of commuting passengers shows strong spatiotemporal regularity. The travel time is concentrated in specific peak hours on weekdays, and the travel OD pairs are concentrated in high-frequency specific lines and station combinations. This regularity provides a reliable basis for demand forecasting. Figure 3 As shown, Figure 3 This is a diagram of the OD distribution of a high-frequency trip of a commuter in City A. Figure 3 The analysis results reveal the main temporal and spatial patterns of commuting passengers.

[0058] like Figure 1 As shown, an embodiment of the present invention provides a method for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics, comprising the following steps: Step S300, extracting travel chain dimension features, time dimension features and space dimension features related to the prediction target according to the travel time and space law.

[0059] In this embodiment, based on the spatiotemporal patterns of commuting passengers, features related to the prediction target are extracted from historical travel behavior data, including travel chain dimension, time dimension, and space dimension features.

[0060] Specifically, in an implementation of this embodiment, step S300 includes the following steps: Step S301, extracting the OD1 data of the same travel order on the same working day, the travel time T1 of the same travel order on the same working day, the most frequent travel time T2 corresponding to the OD1 data, and the most frequent travel time OD2 data corresponding to the travel time T1, to obtain the travel chain dimensional features.

[0061] In this embodiment, the trip chain dimension reflects the behavioral consistency and order regularity of commuting passengers in multiple trips. The specific process of extracting the features of the trip chain dimension includes: (1) Extract the trip OD (origin-destination pairs) of the same working day and the same travel order, denoted as , to capture the common destination patterns of this travel sequence.

[0062] (2) Extract the corresponding travel time, recorded as , analyze the time stability of the travel sequence.

[0063] (3) Statistical correspondence from historical data The highest frequency travel time is denoted as , to reflect its high-frequency temporal pattern.

[0064] (4) Statistics from historical data The corresponding highest frequency travel OD is recorded as , to capture spatial preferences over time.

[0065] Step S302, extracting the travel time of the commuter passengers in the most recent preset trip to obtain the time dimension feature.

[0066] In this embodiment, the characteristics of the time dimension describe the travel regularity of commuting passengers in time. The specific process of extracting the characteristics of the time dimension includes: extracting the travel time data of the last five (i.e., the preset number of times is five) trips of each commuting passenger for analyzing the continuity and regularity in time.

[0067] Step S303, extracting the travel OD data of the commuter passenger corresponding to the most recent preset trip, extracting the highest frequency OD data corresponding to the historical starting point O of each trip that is closest to the end point D of the previous trip, and obtaining the spatial dimension feature.

[0068] In this embodiment, the features of the spatial dimension describe the travel preferences and distribution patterns of commuting passengers in space. The specific process of extracting the features of the spatial dimension includes: (1) The trip OD of each commuter passenger’s most recent five trips (the preset number of trips is 5) is extracted to capture the changing patterns of the start and end points of the trips in the short term.

[0069] (2) Extract the highest frequency OD corresponding to the historical starting point O that is closest to the previous trip destination D to analyze the spatial continuity and preference patterns of the travel chain.

[0070] In this embodiment, in order to ensure the rationality and importance of the extracted features for demand forecasting, it is also necessary to use the conditional probability analysis method to verify the relevance of the above features. The specific process includes: 1. Calculate the conditional probability of the feature: For each feature, calculate the conditional probability value between it and the predicted target; In this embodiment, the calculation formula of conditional probability is expressed as follows:

[0071] The higher the conditional probability value, the stronger the explanatory power and contribution of the feature to the prediction target.

[0072] 2. Screen key travel influencing factors: Based on the calculation results of conditional probability, features with higher conditional probability are selected as model input variables to ensure that the model can capture the core laws and driving factors of travel behavior.

[0073] Through the above steps, the key characteristics of commuter passengers in the travel chain, time and space dimensions were successfully extracted and verified, providing a scientific basis for the construction of subsequent demand forecasting models.

[0074] like Figure 1 As shown, an embodiment of the present invention provides a method for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics, comprising the following steps: Step S400, constructing a feature matrix based on the extracted features, using a distributed gradient enhancement model to train the features in the feature matrix, and predicting and outputting the travel demand of commuting passengers based on the trained travel demand prediction model.

[0075] In this embodiment, based on the extracted spatiotemporal feature data, a machine learning algorithm is used for feature modeling to construct a travel demand forecasting model; that is, based on the extracted spatiotemporal feature data, a feature matrix is ​​constructed, and the XGBoost model (i.e., a distributed gradient boosting model) is used for training and optimization. The trained travel demand forecasting model is used for prediction, and the next travel time and travel OD information of each commuting passenger can be accurately predicted.

[0076] Specifically, in an implementation of this embodiment, step S400 includes the following steps: Step S401, constructing the feature matrix based on the extracted features, and performing one-hot encoding on the trip chain dimension features, time dimension features, and space dimension features in the feature matrix; Step S402, training the travel chain dimension features, time dimension features and space dimension features in the feature matrix based on the distributed gradient enhancement model to obtain the trained travel demand prediction model; In one implementation of this embodiment, step S402 includes the following steps: Step S402a, training the extracted spatiotemporal features based on the distributed gradient enhancement model to build a travel demand prediction model; Step S4022b: Use the K-fold cross-validation method to validate the travel demand forecasting model, and optimize the model parameters according to the validation results to obtain the trained travel demand forecasting model.

[0077] In this embodiment, a machine learning algorithm is used for feature modeling, and the specific process of building a travel demand prediction model includes: 1. Construct a feature matrix: For the extracted travel chain dimension features, time dimension features, and space dimension features, use methods such as One-Hot Encoding to encode the categorical variables and convert them into numerical features suitable for model input; then, construct a feature matrix containing the historical travel behavior characteristics of commuting passengers, and divide the training set and test set into a ratio of 8:2 to ensure the scientific nature of model training and evaluation.

[0078] 2. Model training and optimization: Use the XGBoost algorithm to train the feature matrix, build a travel demand forecasting model, use the K-fold cross-validation method to verify the model performance, optimize the model's hyperparameters (tree depth, learning rate, etc.), and improve the accuracy and robustness of the forecast; select accuracy as the model evaluation indicator to ensure that the model can comprehensively measure the effect of travel demand forecasting.

[0079] 3. Model performance analysis: The experimental verification is conducted using the bus card swiping data from City A from April 24 to May 21, 2018. The data covers approximately 4.53 million passengers and a total of 31.1 million travel records, including route information and OD feature data; The average accuracy of the model during the morning peak (7:00 to 9:00) and evening peak (17:00 to 20:00) on weekdays reached 84.1%, among which the prediction accuracy of the morning peak was the highest, reaching 85.5%; The prediction accuracy during off-peak hours is low, indicating that the model has limited performance in travel scenarios with weak regularity; however, by combining historical travel OD, similar travel day patterns, and high-frequency OD feature combinations, the overall prediction accuracy is further improved, especially in high-frequency travel scenarios.

[0080] Specifically, in an implementation of this embodiment, step S400 further includes the following steps: Step S403: predicting and outputting the travel demand of commuting passengers based on the trained travel demand prediction model.

[0081] In this embodiment, the specific process of performing prediction based on the trained travel demand prediction model includes: 1. Prediction goal: Based on the constructed XGBoost model (trained travel demand prediction model), input the spatiotemporal characteristics of commuting passengers to accurately predict the next travel time and travel OD of each passenger; output the prediction results, including travel time and spatial distribution characteristics, to support the bus company's precise allocation of transportation resources.

[0082] As an example, Figure 4 As shown in the figure, the automatic vehicle positioning AVL data and bus IC card data are input, and then data preprocessing (boarding station matching and alighting station estimation) and travel chain similarity calculation are performed; then, travel characteristics (travel chain dimension characteristics, time dimension characteristics and space dimension characteristics) are extracted; finally, the extracted travel characteristics are input into the XGBoost algorithm model for prediction.

[0083] 2. Analysis of prediction results: (1) Time distribution: From the prediction results, the travel behavior of commuters on weekdays has obvious time regularity, especially in the morning and evening peak hours. The model can effectively capture the characteristics of travel demand during peak hours; the high prediction accuracy during peak hours indicates that commuters have high behavioral repetitiveness and travel regularity during these time periods.

[0084] (2) Spatial distribution: The model accurately predicts the main travel OD pairs of commuters, especially in high-frequency travel scenarios. The prediction in the spatial dimension further verifies the importance of combining features such as the most similar travel day and the historical highest frequency OD. These features play a significant role in capturing spatial patterns.

[0085] This embodiment achieves the following technical effects through the above technical solution: This embodiment can determine the spatiotemporal patterns of commuting passengers by analyzing the historical travel data of public transportation commuting passengers, and by extracting the features of time, space and travel chain dimensions, use the XGBoost algorithm to model the multi-dimensional features, thereby building a machine learning prediction model, which can effectively improve the accuracy of demand forecasting and thus accurately predict the travel needs of commuting passengers.

[0086] Exemplary Devices Based on the above embodiments, the present invention further provides a public transportation commuter travel demand analysis system based on spatiotemporal characteristics, comprising: The dataset acquisition module is used to obtain the historical travel data of bus passengers and construct a personal travel behavior dataset; The travel time-space law analysis module is used to screen commuter passenger groups based on travel frequency, and analyze the travel time-space law based on the passengers' travel time, entropy value of travel stations, and travel chain similarity; A multi-dimensional feature extraction module, used to extract travel chain dimensional features, time dimensional features and space dimensional features related to the prediction target according to the travel time and space law; The prediction and output module is used to construct a feature matrix based on the extracted features, train the features in the feature matrix using a distributed gradient enhancement model, and predict and output the travel demand of commuting passengers based on the trained travel demand prediction model.

[0087] This embodiment achieves the following technical effects through the above technical solution: This embodiment can determine the spatiotemporal patterns of commuting passengers by analyzing the historical travel data of public transportation commuting passengers, and by extracting the features of time, space and travel chain dimensions, use the XGBoost algorithm to model the multi-dimensional features, thereby building a machine learning prediction model, which can effectively improve the accuracy of demand forecasting and thus accurately predict the travel needs of commuting passengers.

[0088] Based on the above embodiment, the present invention further provides a terminal, whose principle block diagram can be as follows: Figure 5 shown.

[0089] The terminal includes: a processor, a memory, an interface, a display screen and a communication module connected through a system bus; wherein the processor of the terminal is used to provide computing and control capabilities; the memory of the terminal includes a storage medium and an internal memory; the storage medium stores an operating system and a computer program; the internal memory provides an environment for the operation of the operating system and the computer program in the storage medium; the interface is used to connect to external devices; the display screen is used to display corresponding information; and the communication module is used to communicate with a cloud server or other devices.

[0090] When the computer program is executed by a processor, it is used to implement the operation of a method for analyzing travel demand of public transportation commuter passengers based on spatiotemporal characteristics.

[0091] It can be understood by those skilled in the art that Figure 5 The principle block diagram shown in the figure is only a block diagram of a partial structure related to the scheme of the present invention, and does not constitute a limitation on the terminal to which the scheme of the present invention is applied. The specific terminal may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0092] In one embodiment, a terminal is provided, which includes: a processor and a memory, wherein the memory stores a public transportation commuter passenger travel demand analysis program based on spatiotemporal characteristics, and the public transportation commuter passenger travel demand analysis program based on spatiotemporal characteristics is used to implement the operation of the above-mentioned public transportation commuter passenger travel demand analysis method based on spatiotemporal characteristics when executed by the processor.

[0093] In one embodiment, a storage medium is provided, wherein the storage medium stores a public transportation commuter travel demand analysis program based on spatiotemporal characteristics, and the public transportation commuter travel demand analysis program based on spatiotemporal characteristics is used to implement the operation of the above-mentioned public transportation commuter travel demand analysis method based on spatiotemporal characteristics when executed by a processor.

[0094] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware through a computer program, and the computer program can be stored in a non-volatile storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention can include non-volatile and volatile memory.

[0095] In summary, the present invention provides a method for analyzing the travel demand of public transportation commuters based on spatiotemporal characteristics, including: obtaining historical travel data of public transportation passengers and constructing a personal travel behavior data set; screening commuter passenger groups based on travel frequency, and analyzing the travel spatiotemporal laws based on the entropy value of the passengers' travel time and travel sites, and the similarity of the travel chain; extracting the travel chain dimension features, time dimension features, and space dimension features related to the prediction target based on the travel spatiotemporal laws; constructing a feature matrix based on the extracted features, training the features in the feature matrix using a distributed gradient enhancement model, and predicting and outputting the travel demand of commuters based on the trained travel demand prediction model. The present invention achieves efficient and accurate prediction of the travel demand of commuters by integrating the multi-dimensional travel characteristics of commuters with machine learning algorithms.

[0096] It should be understood that the application of the present invention is not limited to the above examples. For ordinary technicians in this field, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. A method for analyzing the travel demand of public transportation commuters based on spatiotemporal characteristics, characterized in that: include: Obtain historical travel data of public transport passengers and construct a personal travel behavior dataset; The commuter passenger groups are screened based on the travel frequency, and the travel time and space patterns are analyzed according to the entropy value of the passengers' travel time and travel stations, and the similarity of the travel chain; Extracting travel chain dimension features, time dimension features and space dimension features related to the prediction target according to the travel time and space law; A feature matrix is ​​constructed based on the extracted features, the features in the feature matrix are trained using a distributed gradient enhancement model, and the travel demand of commuting passengers is predicted and output based on the trained travel demand prediction model.

2. The method for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics according to claim 1 is characterized in that: The acquisition of historical travel data of public transport passengers and the construction of a personal travel behavior dataset include: Obtain the boarding station, alighting station, boarding time and alighting time information of the bus passengers to obtain the historical travel data; The personal travel behavior dataset is constructed according to the historical travel data.

3. The method for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics according to claim 1 is characterized in that: The screening of commuter passenger groups based on travel frequency includes: Analyze the daily travel frequency, weekly travel frequency, and repeated travel chain travel frequency of the corresponding passenger based on the personal travel behavior data set; According to the daily travel frequency, the weekly travel frequency and the repeated travel chain travel frequency, commuter passengers are screened from the historical travel data of the bus passengers to determine the commuter passenger group.

4. The method for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics according to claim 1 is characterized in that: The analysis of travel time and space rules based on the entropy value of the passenger's travel time and travel station and the similarity of the travel chain includes: The travel time, entropy values ​​of travel sites and similarity of travel chains on the same weekday corresponding to commuting passengers and non-commuting passengers are calculated respectively, and the spatiotemporal regularity of the travel patterns of the commuting passengers is obtained by comparison.

5. The method for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics according to claim 1 is characterized in that: The extracting of travel chain dimension features, time dimension features and space dimension features related to the prediction target according to the travel time-space law includes: Extract the OD1 data of the same travel order on the same working day, the travel time T1 of the same travel order on the same working day, the most frequent travel time T2 corresponding to the OD1 data, and the most frequent travel time OD2 data corresponding to the travel time T1, to obtain the travel chain dimensional features; Extracting the travel time of the commuter passengers in the most recent preset trip to obtain the time dimension feature; The commuter passenger's travel OD data corresponding to the most recent preset trip is extracted, and the highest frequency OD data corresponding to the historical starting point O that is closest to the previous trip end point D of each trip is extracted to obtain the spatial dimension feature.

6. The method for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics according to claim 1 is characterized in that: The step of constructing a feature matrix based on the extracted features, training the features in the feature matrix using a distributed gradient enhancement model, and predicting and outputting the travel demand of commuting passengers based on the trained travel demand prediction model includes: The feature matrix is ​​constructed based on the extracted features, and the travel chain dimension features, time dimension features, and space dimension features in the feature matrix are uniquely encoded; Based on the distributed gradient enhancement model, the travel chain dimension features, the time dimension features and the space dimension features in the feature matrix are trained to obtain the trained travel demand prediction model; The travel demand of commuting passengers is predicted and output based on the trained travel demand prediction model.

7. The method for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics according to claim 6 is characterized in that: The training of the travel chain dimension features, the time dimension features and the space dimension features in the feature matrix based on the distributed gradient enhancement model includes: Based on the distributed gradient enhancement model, the extracted spatiotemporal features are trained to build a travel demand prediction model; The travel demand forecasting model is verified by using a K-fold cross-validation method, and the model parameters are optimized according to the verification results to obtain the trained travel demand forecasting model.

8. A public transportation passenger travel demand analysis system based on spatiotemporal characteristics, characterized in that: include: The dataset acquisition module is used to obtain the historical travel data of bus passengers and construct a personal travel behavior dataset; The travel time-space law analysis module is used to screen commuter passenger groups based on travel frequency, and analyze the travel time-space law based on the passengers' travel time, entropy value of travel stations, and travel chain similarity; A multi-dimensional feature extraction module, used to extract travel chain dimensional features, time dimensional features and space dimensional features related to the prediction target according to the travel time and space law; The prediction and output module is used to construct a feature matrix based on the extracted features, train the features in the feature matrix using a distributed gradient enhancement model, and predict and output the travel demand of commuting passengers based on the trained travel demand prediction model.

9. A terminal, characterized in that: include: A processor and a memory, wherein the memory stores a program for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics, and when the program for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics is executed by the processor, it is used to implement the operation of the method for analyzing travel demand of public transportation commuters based on spatiotemporal characteristics as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a public transportation commuter passenger travel demand analysis program based on spatiotemporal characteristics, and when the public transportation commuter passenger travel demand analysis program based on spatiotemporal characteristics is executed by a processor, it is used to implement the operation of the public transportation commuter passenger travel demand analysis method based on spatiotemporal characteristics as described in any one of claims 1-7.

Images