Urban public transit hub passenger flow analysis and prediction system using artificial intelligence

Abstract

The invention provides an urban public transit hub passenger flow analysis and prediction system using artificial intelligence. The system comprises a public transit data management module, an urban activity information acquisition module, a weather data acquisition module, a social media data acquisition module, a passenger flow analysis module, a passenger flow prediction module and a model training unit. The bus data management module is used for acquiring bus travel data, clustering and segmenting the bus travel data according to a time sequence and a fixed time interval, storing the clustered and segmented traffic data, and exporting the clustered traffic data into the passenger flow analysis module; the urban activity information acquisition module performs directional crawling on urban activity information, establishes activity influence coefficients according to classification of the urban activity information, calculates correlation coefficients of the urban activity information and public transport hubs according to the urban activity information, and performs grading on the urban activity information according to the correlation coefficients to obtain urban activity grading information. And storing the urban activity grading information.

Description

Technical Field

[0001] The present invention belongs to the field of analysis and prediction, and in particular relates to an urban public transportation hub passenger flow analysis and prediction system using artificial intelligence. Background Art

[0002] At present, in modern urban traffic management, the bus system, as an important public transportation mode, undertakes a large number of passenger transportation tasks. With the acceleration of urbanization, bus hubs, as important nodes of urban public transportation, are particularly important for passenger flow analysis and prediction. Using artificial intelligence technology to analyze and predict passenger flow in urban bus hubs can provide scientific data support for traffic management departments to optimize bus operations, improve service quality and enhance the overall efficiency of urban transportation systems.

[0003] The current related technologies still have some significant defects in practical applications, which limits their effectiveness and reliability in urban public transportation management. First, the existing passenger flow analysis and prediction systems usually rely on historical data for modeling, but the accuracy and completeness of these data are often affected by many factors. For example, there may be omissions or errors in the data collection process, resulting in low quality data used in model training. In addition, historical data may not fully reflect future passenger flow changes, especially when special events (such as holidays, public health emergencies, etc.) occur, the model's predictive ability may be greatly reduced and unable to adapt to dynamically changing passenger flow patterns in a timely manner.

[0004] Many existing artificial intelligence systems are relatively simple in algorithm selection and model construction, lacking diversity and flexibility. Commonly used machine learning models such as linear regression and decision trees can achieve certain results in some scenarios, but when dealing with complex passenger flow data, they may not be able to capture the potential nonlinear relationships and interactive effects in the data, resulting in insufficient accuracy of the prediction results. In addition, existing systems often fail to make full use of multi-source data, such as social media data, weather information, and city event arrangements. These factors have an important impact on passenger flow changes, but are often ignored in current models.

[0005] The current passenger flow prediction system is insufficient in terms of real-time performance. The changes in urban bus passenger flow are highly dynamic, and real-time monitoring and prediction capabilities are essential for traffic management. However, many systems fail to meet real-time requirements in terms of data processing and model reasoning speed, resulting in the inability to timely feedback the prediction results to the traffic management department, affecting the effectiveness of decision-making. At the same time, the existing system needs to be improved in terms of user-friendliness and operability. The implementation of many technical solutions requires professional technicians to perform complex settings and maintenance, while ordinary users often lack the necessary technical knowledge, resulting in a high threshold for using the system. In addition, the system's visualization effects and interactive design are often not intuitive enough, and users may encounter difficulties in understanding and using the system, affecting the popularity and application of the system.

[0006] The current artificial intelligence passenger flow analysis and prediction system is insufficient in terms of model interpretability. Although advanced technologies such as deep learning have performed well in prediction accuracy, their "black box" characteristics make the model's decision-making process difficult to understand, which may cause trouble for traffic managers when making decisions based on model predictions. Models that lack interpretability not only reduce user trust, but may also lead to wrong decisions at critical moments. Therefore, in order to solve the above problems, there is an urgent need for a new type of artificial intelligence urban bus hub passenger flow analysis and prediction system that can combine multi-source data, adopt flexible and diverse models, have real-time processing capabilities, and focus on user experience and model interpretability to improve the scientificity and effectiveness of urban bus management. There is an urgent need for an urban bus hub passenger flow analysis and prediction system that uses artificial intelligence. Summary of the invention

[0007] The present invention proposes an urban bus hub passenger flow analysis and prediction system using artificial intelligence, which solves the problems of single data source and low prediction accuracy in traditional bus passenger flow prediction. By integrating bus travel data, weather information, urban activities and social media data, comprehensive analysis and accurate prediction of passenger flow are achieved.

[0008] The technical solution of the present invention is implemented as follows: a passenger flow analysis and prediction system for urban public transportation hubs using artificial intelligence, comprising a public transportation data management module, an urban activity information acquisition module, a weather data acquisition module, a social media data acquisition module, a passenger flow analysis module, a passenger flow prediction module and a model training unit;

[0009] The bus data management module is used to obtain bus travel data, cluster and segment the bus travel data in chronological order at fixed time intervals, store the clustered and segmented traffic data, and then export the clustered traffic data to the passenger flow analysis module;

[0010] The city activity information acquisition module performs targeted crawling of city activity information, establishes activity influence coefficients according to the classification of city activity information, calculates the correlation coefficient between city activity information and public transportation hubs according to the city activity information, classifies the city activity information according to the correlation coefficient to obtain city activity classification information, and stores the city activity classification information; then, the city activity classification information is exported to the passenger flow analysis module;

[0011] The weather data acquisition module collects historical weather data, extracts weather characteristics based on the historical weather data and stores them; then the weather characteristics are exported to the passenger flow analysis module

[0012] The social media data acquisition module calls the social media data in real time, extracts the correlation coefficient between the social media data and the public transportation hub, classifies the social media data to obtain social media data classification information, and stores the social media data classification information; then, the social media data classification information is exported to the passenger flow analysis module;

[0013] The passenger flow analysis module performs neural network model training according to the clustered and segmented traffic data, urban activity classification information, social media data classification information and weather characteristics to obtain a trained passenger flow analysis model; the passenger flow analysis module is built with a data preprocessing unit, through which the clustered traffic data, urban activity classification information, social media data classification information and weather characteristics are marked and analyzed and preprocessed data are obtained, and then the preprocessed data is sent to the model training unit;

[0014] The passenger flow prediction module includes: a first data processing unit and a second data processing unit, wherein the first data processing unit is used to perform data preprocessing on real-time collected traffic data, urban activity classification information, social media data classification information and weather characteristics to obtain preprocessed data; wherein the second data processing unit is used to input the preprocessed data into the trained passenger flow analysis model to obtain passenger flow prediction data of the bus hub within a preset time interval in the future;

[0015] The model training unit is used to input the preprocessed data into the neural network model for training, obtain the trained passenger flow analysis model, and feed the passenger flow analysis model back to the passenger flow analysis module.

[0016] Traditional passenger flow prediction systems usually rely on a single data source, such as relying solely on bus travel data or historical passenger flow data for analysis. This approach is easily limited by data quality and coverage, resulting in inaccurate prediction results. This system integrates multiple data sources, including bus travel data, urban activity information, weather data, and social media data, to form a comprehensive passenger flow analysis framework. This multi-dimensional data integration method enables the system to more comprehensively reflect the various factors that affect passenger flow, thereby improving the accuracy and reliability of the prediction.

[0017] The modular design of the system enhances the collaborative working ability between various parts. The passenger flow analysis of existing technologies is often a linear process, lacking effective connection and data interaction between modules, resulting in the phenomenon of information islands. However, this system designs multifunctional modules such as bus data management module, urban activity information acquisition module, weather data acquisition module, and social media data acquisition module, so that data and information can be shared in real time between modules. This modular design not only improves the efficiency of data processing, but also enhances the flexibility of the system, allowing the integration and analysis of different data sources to be carried out more quickly.

[0018] This system uses a neural network model to analyze and predict passenger flow, which is also a technical innovation. Traditional passenger flow prediction methods usually rely on simple statistical models or linear regression, which are difficult to capture complex nonlinear relationships. The neural network model can automatically extract features from a large amount of historical data through deep learning technology and identify complex patterns and trends. This application of deep learning enables the system to provide more accurate prediction results when facing complex passenger flow dynamics, significantly improving the accuracy of predictions.

[0019] The real-time and dynamic nature of the system is also a major advantage. The passenger flow prediction in the existing technology often relies on static data and lacks the ability to respond to real-time changes. However, this system realizes real-time monitoring and dynamic prediction of passenger flow changes through social media data acquisition modules and real-time traffic data collection modules. This real-time nature enables the system to quickly respond to emergencies, such as urban activities, weather changes, etc., and adjust the prediction results in time, improving the practicality and effectiveness of the system.

[0020] In the process of data processing and analysis, the system uses a data preprocessing unit to perform labeling analysis and feature extraction on various types of data. This data preprocessing process not only improves data quality, but also provides a more reliable foundation for subsequent model training. Traditional systems often ignore the importance of data preprocessing, which affects the results of subsequent analysis. This system ensures the effectiveness and accuracy of the analysis model through a sophisticated data processing process.

[0021] As a preferred implementation, the bus data management module collects bus travel data, cleans and integrates the bus travel data; the bus travel data includes bus schedule information and platform passenger flow information of each bus hub; the pre-processed bus travel data is obtained through a data mining unit, the bus travel data is clustered and segmented in chronological order at fixed time intervals, the data in the same time period is divided according to vehicles according to vehicle status information, and the clustered and segmented traffic data is obtained; the data mining unit is built with a first data cleaning subunit and a data division subunit, the first data cleaning subunit pre-processes the bus travel data, the bus travel data is clustered and segmented in chronological order at fixed time intervals; the data division subunit is used to obtain the clustered and segmented data, the data in the same time period is divided according to vehicles according to vehicle status information, and the platform passenger flow corresponding to the time period and the number of vehicles is obtained; the bus data management module is also provided with a database unit, the database unit is used to store the clustered and segmented traffic data.

[0022] As a preferred embodiment, the neural network model of the passenger flow analysis module includes four neural network models, which are respectively recorded as passenger flow analysis neural network model A, passenger flow analysis neural network model B, passenger flow analysis neural network model C and passenger flow analysis neural network model D; the passenger flow prediction module includes four passenger flow prediction sub-modules, which generate corresponding passenger flow prediction sub-modules according to passenger flow analysis neural network model A, passenger flow analysis neural network model B, passenger flow analysis neural network model C and passenger flow analysis neural network model D, respectively, wherein the passenger flow prediction sub-module is used to input real-time collected traffic data, urban activity classification information, social media classification data and weather characteristics into the corresponding trained passenger flow analysis model to perform passenger flow prediction, and obtain multi-dimensional passenger flow prediction data for the bus hub within a preset time interval in the future.

[0023] As a preferred implementation, the passenger flow analysis neural network model A is trained based on clustered traffic data to obtain a trained passenger flow analysis neural network model A.

[0024] As a preferred implementation, the passenger flow analysis neural network model B is trained according to the city activity classification information to obtain the trained passenger flow analysis neural network model B.

[0025] As a preferred implementation, the passenger flow analysis neural network model C is trained according to social media data classification information to obtain a trained passenger flow analysis neural network model C.

[0026] As a preferred implementation, the passenger flow analysis neural network model D is trained according to weather characteristics to obtain a trained passenger flow analysis neural network model D.

[0027] After adopting the above technical solution, the beneficial effect of the present invention is that by integrating multiple data sources, the system can comprehensively consider multiple influencing factors and conduct a comprehensive analysis of passenger flow changes. This all-round data processing method makes the prediction results more accurate and effectively reduces the prediction errors caused by the singleness of the data, thereby providing a more reliable decision-making basis for the public transportation management department.

[0028] The system's real-time data processing capabilities significantly improve its ability to respond to emergencies. Urban activities, bad weather and other factors often have a significant impact on public transportation passenger flow, and traditional systems may not be able to capture these changes in a timely manner. By calling social media data and weather information in real time, this system can quickly identify and predict the impact of these emergencies on passenger flow, so that public transportation operations can be adjusted in a timely manner to ensure the continuity and stability of services. This dynamic response capability not only improves users' travel experience, but also enhances the adaptability of public transportation systems.

[0029] The modular design of the system makes the collaboration between various functional modules more efficient. By modularizing functions such as bus data management, urban activity acquisition, and weather data collection, the system can quickly integrate various types of information to form a complete passenger flow analysis system. This efficient collaboration mode makes the conversion and analysis between different data sources smoother and improves the overall system efficiency. The system's deep learning model provides strong technical support for passenger flow prediction. Through the application of neural network models, the system can automatically learn and extract features from massive data and identify complex passenger flow patterns. The application of this technology enables the system to provide more accurate prediction results when facing complex traffic environments, reduces the need for manual intervention, and improves the level of automation of analysis. The implementation of this system will promote the development of smart cities. By improving the accuracy and real-time performance of bus passenger flow prediction, the system can optimize bus scheduling and resource allocation and improve the operational efficiency of public transportation. This will not only help alleviate urban traffic congestion, but also increase the attractiveness of public transportation and encourage more citizens to choose public transportation, thereby promoting the development of sustainable urban transportation. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] 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 these drawings without paying creative labor.

[0031] Figure 1 It is a flow chart of the system of the present invention. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] Example:

[0034] like Figure 1 As shown, a passenger flow analysis and prediction system for urban public transportation hubs using artificial intelligence includes a public transportation data management module, an urban activity information acquisition module, a weather data acquisition module, a social media data acquisition module, a passenger flow analysis module, a passenger flow prediction module, and a model training unit;

[0035] The bus data management module is used to obtain bus travel data, cluster and segment the bus travel data in chronological order at fixed time intervals, store the clustered and segmented traffic data, and then export the clustered traffic data to the passenger flow analysis module;

[0036] The city activity information acquisition module performs targeted crawling of city activity information, establishes activity influence coefficients according to the classification of city activity information, calculates the correlation coefficient between city activity information and public transportation hubs according to the city activity information, classifies the city activity information according to the correlation coefficient to obtain city activity classification information, and stores the city activity classification information; then, the city activity classification information is exported to the passenger flow analysis module;

[0037] The weather data acquisition module collects historical weather data, extracts weather characteristics based on the historical weather data and stores them; then the weather characteristics are exported to the passenger flow analysis module

[0038] The social media data acquisition module calls the social media data in real time, extracts the correlation coefficient between the social media data and the public transportation hub, classifies the social media data to obtain social media data classification information, and stores the social media data classification information; then, the social media data classification information is exported to the passenger flow analysis module;

[0039] The passenger flow analysis module performs neural network model training according to the clustered and segmented traffic data, urban activity classification information, social media data classification information and weather characteristics to obtain a trained passenger flow analysis model; the passenger flow analysis module is built with a data preprocessing unit, through which the clustered traffic data, urban activity classification information, social media data classification information and weather characteristics are marked and analyzed and preprocessed data are obtained, and then the preprocessed data is sent to the model training unit;

[0040] The passenger flow prediction module includes: a first data processing unit and a second data processing unit, wherein the first data processing unit is used to perform data preprocessing on real-time collected traffic data, urban activity classification information, social media data classification information and weather characteristics to obtain preprocessed data; wherein the second data processing unit is used to input the preprocessed data into the trained passenger flow analysis model to obtain passenger flow prediction data of the bus hub within a preset time interval in the future;

[0041] The model training unit is used to input the preprocessed data into the neural network model for training, obtain the trained passenger flow analysis model, and feed the passenger flow analysis model back to the passenger flow analysis module.

[0042] This technical solution adopts an urban bus hub passenger flow analysis and prediction system using artificial intelligence. The system integrates multiple data sources and realizes the analysis and prediction of passenger flow through a series of modules. Its working principle and workflow can be divided into multiple steps, and the specific operations are as follows.

[0043] The core of the system is the bus data management module, which is responsible for acquiring and processing bus travel data. This module collects bus-related travel data from different data sources, including the time, location, vehicle number, and other information of passengers getting on and off the bus. The acquired data will be processed in chronological order and clustered and segmented according to preset fixed time intervals. This clustering process helps to classify the data by time period, making subsequent analysis more organized. After processing, the clustered and segmented traffic data will be stored and exported to the passenger flow analysis module to become the basic data for subsequent analysis.

[0044] The city activity information acquisition module is responsible for crawling information about various activities taking place in the city. This module will obtain city activity information in a targeted manner, including festivals, exhibitions, competitions, etc., and then establish the corresponding activity impact coefficient based on the nature and influence of the activity. By calculating the correlation coefficient between city activity information and bus hubs, the module will classify the activity information to obtain city activity classification information. This information will be stored and exported to the passenger flow analysis module to help analyze the impact of activities on passenger flow.

[0045] The weather data acquisition module is responsible for collecting historical weather data, which includes weather characteristics such as temperature, rainfall, and wind speed. By analyzing historical weather data, the module extracts relevant weather characteristics, stores these characteristics, and then exports them to the passenger flow analysis module. The introduction of this data can help the model understand the impact of weather on passenger flow, thereby improving the accuracy of predictions.

[0046] The social media data acquisition module calls the data of social media platforms in real time and analyzes the correlation between these data and bus hubs. By extracting social media data, the module calculates the correlation coefficient and classifies it to obtain social media data classification information. This information will also be stored and exported to the passenger flow analysis module, providing real-time social trends and public sentiment feedback for passenger flow analysis.

[0047] In the passenger flow analysis module, the system combines clustered and segmented traffic data, urban activity classification information, social media data classification information, and weather characteristics, and trains them through a neural network model. This module has a built-in data preprocessing unit, which first performs labeling analysis on the clustered traffic data, urban activity classification information, social media data classification information, and weather characteristics, obtains the preprocessed data, and sends it to the model training unit. Through this training process, the system can build an accurate passenger flow analysis model.

[0048] The passenger flow prediction module consists of a first data processing unit and a second data processing unit. The task of the first data processing unit is to preprocess the real-time collected traffic data, urban activity classification information, social media data classification information and weather characteristics. This process ensures that the input data meets the requirements of model training and prediction, and removes noise and irrelevant information. After completing the preprocessing, the second data processing unit inputs this data into the trained passenger flow analysis model to generate passenger flow prediction data for the bus hub within a preset time interval in the future.

[0049] The workflow of the entire system is a cyclic process. The model training unit inputs the pre-processed data into the neural network model for training to obtain an updated passenger flow analysis model. This trained model will be fed back to the passenger flow analysis module to form a closed loop, allowing the model to continuously learn and optimize to adapt to the dynamically changing urban bus passenger flow situation.

[0050] This system integrates public transportation data, urban activity information, weather data and social media data, and uses advanced artificial intelligence technology for data analysis and prediction, which can effectively improve the passenger flow management efficiency and service quality of urban public transportation hubs. Through the collaboration of these modules, the system realizes comprehensive analysis and accurate prediction of passenger flow, providing strong data support and decision-making basis for urban traffic management.

[0051] The bus data management module collects bus travel data and cleans and integrates the bus travel data; the bus travel data includes bus schedule information and platform passenger flow information of each bus hub; the pre-processed bus travel data is obtained through a data mining unit, the bus travel data is clustered and segmented in chronological order at fixed time intervals, and the data in the same time period is divided according to vehicles according to vehicle status information to obtain clustered and segmented traffic data; the data mining unit is built with a first data cleaning subunit and a data division subunit, the first data cleaning subunit pre-processes the bus travel data, and the bus travel data is clustered and segmented in chronological order at fixed time intervals; the data division subunit is used to obtain the clustered and segmented data, and the data in the same time period is divided according to vehicles according to vehicle status information to obtain the platform passenger flow corresponding to the time period and the number of vehicles; the bus data management module is also provided with a database unit, and the database unit is used to store the clustered and segmented traffic data.

[0052] In this technical solution, the bus data management module significantly improves the efficiency and accuracy of data processing by cleaning and integrating bus travel data. This module focuses on collecting bus travel data including bus schedule information and platform passenger flow information of each bus hub. In the prior art, data acquisition and processing often lack systematicity and standardization, resulting in low data quality and affecting the effect of subsequent analysis. However, this solution pre-processes the bus travel data through a data mining unit to ensure the integrity and consistency of the data. Specifically, the first data cleaning subunit cleans the original data to remove redundant and erroneous information to ensure the accuracy of the data. Then, the data partitioning subunit clusters and segments the cleaned data according to time sequence and fixed time intervals, and divides the data in the same time period according to the vehicle status information to obtain the platform passenger flow corresponding to the time period and the number of vehicles. This process not only improves the efficiency of data processing, but also makes the data more structured, which is conducive to subsequent passenger flow analysis. In contrast, the prior art often uses a single time tag or a simple clustering method, lacks a detailed division of vehicle status, and results in insufficient accuracy in passenger flow analysis. In addition, a database unit is also set up in the bus data management module to store clustered and segmented traffic data. This modular design makes data storage and management more efficient and facilitates subsequent data calls and analysis.

[0053] The neural network model of the passenger flow analysis module includes four neural network models, which are respectively recorded as passenger flow analysis neural network model A, passenger flow analysis neural network model B, passenger flow analysis neural network model C and passenger flow analysis neural network model D; the passenger flow prediction module includes four passenger flow prediction sub-modules, which generate corresponding passenger flow prediction sub-modules according to passenger flow analysis neural network model A, passenger flow analysis neural network model B, passenger flow analysis neural network model C and passenger flow analysis neural network model D, respectively, wherein the passenger flow prediction sub-module is used to input real-time collected traffic data, urban activity classification information, social media classification data and weather characteristics into the corresponding trained passenger flow analysis model to perform passenger flow prediction, and obtain multi-dimensional passenger flow prediction data of the bus hub within a preset time interval in the future.

[0054] In the passenger flow analysis module, this solution introduces four different neural network models, namely passenger flow analysis neural network models A, B, C and D. This design idea is in sharp contrast to the single model approach of the existing technology. The existing technology usually relies on a single neural network model to process all data, which may not be able to fully explore the potential information of different types of data. However, this solution can better capture the complex relationship between data by building multiple models and training them for different data sources. For example, passenger flow analysis neural network model A is mainly trained based on clustered traffic data, while model B focuses on urban activity classification information, model C is based on social media data classification information, and model D is trained based on weather characteristics. This multi-model design not only improves the accuracy of analysis, but also enhances the system's adaptability to various influencing factors. By inputting real-time collected traffic data, urban activity classification information, social media classification data and weather characteristics into the corresponding trained passenger flow analysis model, the system can generate multi-dimensional passenger flow prediction data. This multi-dimensional prediction capability is lacking in the existing technology, which can more comprehensively reflect the passenger flow dynamics of the bus hub and provide more accurate data support for urban traffic management.

[0055] The passenger flow analysis neural network model A is trained based on the clustered traffic data to obtain the trained passenger flow analysis neural network model A. During the training process of the passenger flow analysis neural network model A, the model focuses on the clustered traffic data. Through in-depth analysis of such data, the model can learn the laws and trends of passenger flow changes, thereby generating more accurate passenger flow forecasts. This process is in contrast to the single data processing model of the prior art, which often cannot make full use of the rich traffic data, resulting in limitations in the prediction results. This solution, by specifically training on clustered traffic data, ensures that the model can fully tap the potential information in the data and improves the accuracy of the prediction.

[0056] The passenger flow analysis neural network model B is trained according to the urban activity classification information to obtain the trained passenger flow analysis neural network model B. Similarly, the passenger flow analysis neural network model B is mainly trained based on the urban activity classification information. This training method focusing on a specific data source can effectively capture the impact of urban activities on public transportation passenger flow. The existing technology often ignores this important factor, resulting in the one-sidedness of passenger flow prediction. This solution introduces relevant data on urban activities, so that the model can more comprehensively consider various factors affecting passenger flow, thereby improving the scientificity and accuracy of the prediction.

[0057] The passenger flow analysis neural network model C is trained according to the classification information of social media data to obtain the trained passenger flow analysis neural network model C. The passenger flow analysis neural network model C is trained according to the classification information of social media data. This design highlights the impact of social media on public travel decisions. Existing technologies often fail to fully consider the real-time and dynamic nature of social media. This solution, by incorporating social media data into the analysis, can timely reflect public sentiment and trends, and provide a more real-time basis for passenger flow prediction. This method not only improves the accuracy of passenger flow prediction, but also enhances the adaptability of the system.

[0058] The passenger flow analysis neural network model D is trained according to weather characteristics to obtain the trained passenger flow analysis neural network model D. The passenger flow analysis neural network model D focuses on the training of weather characteristics. The impact of weather factors on public transportation passenger flow is obvious. The existing technology often fails to fully integrate weather data into the passenger flow prediction model, resulting in deviations in the prediction results. However, this solution enables the model to better capture the impact of weather changes on passenger flow by training specifically for weather characteristics, thereby improving the accuracy and reliability of the prediction.

[0059] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An urban public transport hub passenger flow analysis and prediction system using artificial intelligence, characterized in that: It includes bus data management module, urban activity information acquisition module, weather data acquisition module, social media data acquisition module, passenger flow analysis module, passenger flow prediction module and model training unit; The bus data management module is used to obtain bus travel data, cluster and segment the bus travel data in chronological order at fixed time intervals, store the clustered and segmented traffic data, and then export the clustered traffic data to the passenger flow analysis module; The city activity information acquisition module performs targeted crawling of city activity information, establishes activity influence coefficients according to the classification of city activity information, calculates the correlation coefficient between city activity information and public transportation hubs according to the city activity information, classifies the city activity information according to the correlation coefficient to obtain city activity classification information, and stores the city activity classification information; then, the city activity classification information is exported to the passenger flow analysis module; The weather data acquisition module collects historical weather data, extracts weather features based on the historical weather data and stores them; The weather characteristics are then exported to the passenger flow analysis module; The social media data acquisition module calls the social media data in real time, extracts the correlation coefficient between the social media data and the public transportation hub, classifies the social media data to obtain social media data classification information, and stores the social media data classification information; Then export the social media data classification information to the customer flow analysis module; The passenger flow analysis module performs neural network model training according to the clustered and segmented traffic data, urban activity classification information, social media data classification information and weather characteristics to obtain a trained passenger flow analysis model; the passenger flow analysis module is built with a data preprocessing unit, through which the clustered traffic data, urban activity classification information, social media data classification information and weather characteristics are marked and analyzed and preprocessed data are obtained, and then the preprocessed data is sent to the model training unit; The passenger flow prediction module includes: a first data processing unit and a second data processing unit, wherein the first data processing unit is used to perform data preprocessing on real-time collected traffic data, urban activity classification information, social media data classification information and weather characteristics to obtain preprocessed data; wherein the second data processing unit is used to input the preprocessed data into the trained passenger flow analysis model to obtain passenger flow prediction data of the bus hub within a preset time interval in the future; The model training unit is used to input the preprocessed data into the neural network model for training, obtain the trained passenger flow analysis model, and feed the passenger flow analysis model back to the passenger flow analysis module.

2. The urban public transport hub passenger flow analysis and prediction system using artificial intelligence as claimed in claim 1, characterized in that: The bus data management module collects bus travel data, cleans and integrates the bus travel data; the bus travel data includes bus schedule information and platform passenger flow information of each bus hub; The pre-processed bus travel data is obtained through a data mining unit, and the bus travel data is clustered and segmented in chronological order at fixed time intervals. The data in the same time period is divided according to vehicles according to vehicle status information to obtain clustered and segmented traffic data; the data mining unit is built with a first data cleaning subunit and a data division subunit, the first data cleaning subunit pre-processes the bus travel data, and clusters and segments the bus travel data in chronological order at fixed time intervals; the data division subunit is used to obtain the clustered and segmented data, and divides the data in the same time period according to vehicles according to vehicle status information to obtain the platform passenger flow corresponding to the time period and the number of vehicles; the bus data management module is also provided with a database unit, and the database unit is used to store the clustered and segmented traffic data.

3. The urban public transport hub passenger flow analysis and prediction system using artificial intelligence as claimed in claim 1, characterized in that: The neural network model of the passenger flow analysis module includes four neural network models, which are respectively recorded as passenger flow analysis neural network model A, passenger flow analysis neural network model B, passenger flow analysis neural network model C and passenger flow analysis neural network model D; the passenger flow prediction module includes four passenger flow prediction sub-modules, which generate corresponding passenger flow prediction sub-modules according to passenger flow analysis neural network model A, passenger flow analysis neural network model B, passenger flow analysis neural network model C and passenger flow analysis neural network model D, respectively, wherein the passenger flow prediction sub-module is used to input real-time collected traffic data, urban activity classification information, social media classification data and weather characteristics into the corresponding trained passenger flow analysis model to perform passenger flow prediction, and obtain multi-dimensional passenger flow prediction data of the bus hub within a preset time interval in the future.

4. The urban public transport hub passenger flow analysis and prediction system using artificial intelligence as claimed in claim 3, characterized in that: The passenger flow analysis neural network model A is trained according to the clustered traffic data to obtain the trained passenger flow analysis neural network model A.

5. The urban public transport hub passenger flow analysis and prediction system using artificial intelligence as claimed in claim 3, characterized in that: The passenger flow analysis neural network model B is trained according to the city activity classification information to obtain the trained passenger flow analysis neural network model B.

6. The urban public transport hub passenger flow analysis and prediction system using artificial intelligence as claimed in claim 3, characterized in that: The passenger flow analysis neural network model C is trained according to the social media data classification information to obtain the trained passenger flow analysis neural network model C.

7. The urban public transportation hub passenger flow analysis and prediction system using artificial intelligence as claimed in claim 3, characterized in that: The passenger flow analysis neural network model D is trained according to weather characteristics to obtain a trained passenger flow analysis neural network model D.

Images