Public transport means management method and device based on multi-source data and medium

Abstract

The invention discloses a public transport means management method and device based on multi-source data and a medium. Comprising the following steps: acquiring positioning data, get-on and get-off passenger number data, station waiting passenger number data and meteorological data of a plurality of public transport means in real time; according to the meteorological data, the time characteristics and the station surrounding event characteristics, carrying out nonlinear fitting on the getting-on and getting-off passenger number data and the station waiting passenger number data to obtain a passenger flow density value of each station; dividing a plurality of line section intervals through the road network data, and determining line section passenger flow upper limit data according to the passenger flow density value of each station included in the line section intervals; and adjusting the scheduling information of the public transport means according to the line section passenger flow upper limit data. The passenger flow information is determined by fusing multi-source data, so that the accuracy of the passenger flow information is improved; line section interval division is performed through road network data, and public transport means scheduling is performed according to the passenger flow data of the line section interval, so that the scheduling accuracy is improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent transportation technology, and in particular to a method, apparatus and medium for managing public transportation vehicles based on multi-source data. Background Technology

[0002] With the acceleration of urbanization and the continuous growth of residents' travel demands, the core position of public transportation systems in urban transportation is becoming increasingly prominent. Traditional public transportation management mainly relies on fixed timetables, manual dispatching, and limited onboard equipment data, making it difficult to dynamically respond to real-time changes in passenger flow, road congestion, and emergencies. This leads to problems such as uneven capacity allocation, long passenger waiting times, and high vehicle empty-running rates, seriously affecting the service efficiency and user experience of public transportation.

[0003] While some current intelligent public transportation systems have introduced some data analysis functions, they mostly focus on a single data source (such as estimating passenger flow based solely on card swipe data) or only support post-event statistical analysis. They lack an integrated management architecture oriented towards the closed loop of "perception-prediction-decision-execution" and cannot support dynamic scheduling and resource allocation with high timeliness and robustness. Summary of the Invention

[0004] This invention provides a public transportation management method, device, and medium based on multi-source data to solve the problems of inaccurate public transportation scheduling schemes, which lead to the inability to meet passenger travel needs and low resource utilization.

[0005] According to one aspect of the present invention, a public transportation vehicle management method based on multi-source data is provided, comprising:

[0006] Real-time collection of location data, passenger boarding and alighting data, station waiting data, and weather data from multiple public transportation vehicles;

[0007] Based on the meteorological data, time characteristics, and event characteristics around the station, the passenger boarding and alighting data and the waiting number data at the station are nonlinearly fitted to obtain the passenger flow density value of each station.

[0008] Multiple route section intervals are delineated using road network data, and the upper limit data of the route section passenger flow is determined based on the passenger flow density value of each station included in the route section interval.

[0009] The scheduling information of the public transportation vehicles is adjusted based on the passenger flow limit data of the aforementioned route section.

[0010] According to another aspect of the present invention, a public transportation vehicle management device based on multi-source data is provided, comprising:

[0011] The multi-source data acquisition module is used to collect real-time location data, passenger boarding and alighting data, station waiting data, and meteorological data from multiple public transportation vehicles.

[0012] The station passenger flow density value determination module is used to perform nonlinear fitting on the passenger boarding and alighting data and the station waiting number data based on the meteorological data, time characteristics, and station surrounding event characteristics to obtain the passenger flow density value of each station.

[0013] The module for determining the upper limit of passenger flow for a line section is used to delineate multiple line section intervals through road network data and determine the upper limit of passenger flow for the line section based on the passenger flow density value of each station included in the line section interval.

[0014] The scheduling adjustment module is used to adjust the scheduling information of the public transportation vehicles based on the passenger flow limit data of the line section.

[0015] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0016] At least one processor; and

[0017] A memory communicatively connected to the at least one processor; wherein,

[0018] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the public transportation management method based on multi-source data according to any embodiment of the present invention.

[0019] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the public transportation vehicle management method based on multi-source data as described in any embodiment of the present invention.

[0020] According to another aspect of this application, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the public transportation management method based on multi-source data as described in any embodiment of this application.

[0021] The technical solution of this invention improves the accuracy of passenger flow information determination by integrating multi-source data to determine passenger flow information of public transportation vehicles; furthermore, it divides the route section into segments using road network data, analyzes the passenger flow data within the segment, and then schedules public transportation vehicles based on the passenger flow data of the segment, thereby improving the accuracy of scheduling and avoiding errors caused by scheduling based on information from a single station.

[0022] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart of a public transportation vehicle management method based on multi-source data according to an embodiment of the present invention;

[0025] Figure 2 This is a flowchart of another public transportation vehicle management method based on multi-source data provided by an embodiment of the present invention;

[0026] Figure 3 This is a flowchart of another public transportation vehicle management method based on multi-source data provided by an embodiment of the present invention;

[0027] Figure 4 This is a schematic diagram of the architecture of a public transportation vehicle management system based on multi-source data according to an embodiment of the present invention;

[0028] Figure 5 This is a schematic diagram of the structure of a multi-source data access module provided in an embodiment of the present invention;

[0029] Figure 6 This is a schematic diagram of the data processing process of the multi-source data access module provided in an embodiment of the present invention;

[0030] Figure 7 This is a schematic diagram of the structure of the core module of the data platform provided in an embodiment of the present invention;

[0031] Figure 8 This is a schematic diagram of the data processing process of the dynamic passenger flow processing module, cross-section analysis module, and intelligent scheduling module provided in an embodiment of the present invention.

[0032] Figure 9 This is a schematic diagram of the data processing process of the core module of the data platform provided in an embodiment of the present invention;

[0033] Figure 10 This is a runtime sequence diagram of a public transportation vehicle management system based on multi-source data provided according to an embodiment of the present invention;

[0034] Figure 11 This is a schematic diagram of a public transportation vehicle management device based on multi-source data according to an embodiment of the present invention;

[0035] Figure 12 This is a schematic diagram of the structure of an electronic device that implements the public transportation vehicle management method based on multi-source data according to an embodiment of the present invention. Detailed Implementation

[0036] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0037] It should be noted that the terms "candidate," "target," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0038] Figure 1 This invention provides a flowchart of a public transportation vehicle management method based on multi-source data. This embodiment is applicable to optimizing public transportation vehicle scheduling schemes. The method can be executed by a public transportation vehicle management device based on multi-source data, which can be implemented in hardware and / or software and configured on a server. Figure 1 As shown, the method includes:

[0039] S110 collects real-time location data, passenger boarding and alighting data, station waiting data, and weather data from multiple public transportation vehicles.

[0040] Public transportation vehicles can be those used within public transportation networks such as buses or subways.

[0041] Location data for public transportation vehicles can be obtained from GPS positioning devices deployed on the vehicles, and can be used to reflect the real-time movement and location information of public transportation vehicles.

[0042] Passenger boarding and alighting data for public transportation vehicles are used to reflect the number of passengers boarding and alighting after the vehicle arrives at a station. For example, corresponding sensors can be deployed at the doors of public transportation vehicles. For instance, a pair of infrared or laser transmitting and receiving devices can be installed above the doors of buses to count passengers based on the transmitted and received information. Alternatively, visual cameras can be deployed at the doors, and the number of passengers boarding and alighting can be determined by analyzing the image information collected by the cameras.

[0043] Public transportation station waiting data is used to reflect the number of people waiting to take public transportation at each station. For example, images of the station area are acquired by visual cameras deployed at each station, and the number of people waiting in the images of the station area is analyzed and counted.

[0044] Meteorological data is used to reflect weather conditions and can be determined based on officially released meteorological data.

[0045] For example, location data acquisition sources and passenger boarding / alighting data acquisition sources are deployed on public transportation vehicles, and waiting passenger data acquisition sources are deployed at each stop along the route of the public transportation vehicle. A communication link is established between each acquisition source and a backend server so that the real-time collected data can be sent to the backend server through the communication link. At the same time, the backend server obtains real-time weather data from an official platform. For example, the location data acquisition source can be a GPS positioning device, the passenger boarding / alighting data acquisition source can be an infrared or laser transmitting and receiving device, and the waiting passenger data acquisition source can be a visual camera, etc.

[0046] In a feasible embodiment, before performing nonlinear fitting on the passenger boarding / alighting data and the station waiting area data based on meteorological data, time characteristics, and station surrounding event characteristics, the method further includes:

[0047] The positioning data is filtered using a Kalman filter, and the current trajectory information is determined based on the filtered positioning data.

[0048] The current trajectory information is corrected based on the matching results between historical trajectory information and current trajectory information;

[0049] Based on the timestamp synchronization results of location data and passenger boarding / alighting data, the passenger boarding / alighting data is filtered.

[0050] Before performing subsequent calculations based on multi-source data, it is necessary to clean the multi-source data to improve the accuracy of the results. This specifically includes cleaning the location data and cleaning the passenger boarding and alighting data.

[0051] For example, due to satellite signal obstruction and multipath effects, vehicle GPS positioning data generates random errors, manifesting as positioning point errors. Therefore, a Kalman filter is first used to filter the collected positioning data, resulting in filtered positioning data. The core advantage of the Kalman filter algorithm lies in its ability to handle data containing random noise. Through an iterative prediction-update process, it separates the effective components from noise in the signal, thereby reducing errors. The current trajectory information for each public transportation vehicle is then generated based on the filtered positioning data. For instance, after acquiring positioning data from the vehicle GPS, fluctuating points are filtered out, retaining the effective trajectories within the speed range of 0-60 km / h. The positioning error is then corrected using the Kalman filter algorithm to obtain the current trajectory information.

[0052] Furthermore, due to the continuous and constrained movement of public transportation vehicles—meaning they must follow predetermined routes—severe deviations in the current trajectory can be corrected based on historical trajectory information. For example, the current deviation between the current and historical trajectory information can be determined. If the current deviation exceeds a threshold, the current trajectory can be corrected based on historical trajectory information, such as replacing the portion of the trajectory with the corresponding part from the historical trajectory. Kalman filtering, by establishing a vehicle motion state equation and combining it with GPS measurement equations, can predict the current position based on historical trajectory and dynamically adjust the prediction results based on new measurements, making the corrected trajectory more consistent with the actual vehicle motion. In practical applications, for example, when GPS suddenly outputs a drift point exceeding the reasonable speed range, the algorithm will use state prediction and measurement residual analysis to determine that the point is an outlier and correct it, avoiding trajectory distortion.

[0053] Because public transportation vehicles may pick up and drop off passengers at non-stop locations, such as due to driver violations, this is considered an abnormal situation. Therefore, the passenger pick-up and drop-off data in such cases needs to be deleted to ensure the analysis of passenger pick-up and drop-off data under normal circumstances. For example, a first timestamp for location data and a second timestamp for passenger pick-up and drop-off data are determined. The location information corresponding to the passenger pick-up and drop-off data is determined based on the matching of the first and second timestamps. It is then determined whether this location information corresponds to any stop location. If not, the passenger pick-up and drop-off data corresponding to the second timestamp is deleted; otherwise, it is retained. For example, by synchronizing with a GPS trajectory cleaning unit via timestamps, the door opening and closing signals are matched with stop locations to eliminate erroneous statistical data caused by vehicles not stopping at designated locations.

[0054] This embodiment improves the accuracy of the collected data by cleaning the location data and the number of passengers getting on and off, thereby improving the accuracy of the subsequent determination of relevant parameters.

[0055] S120. Based on meteorological data, time characteristics, and event characteristics around the station, nonlinear fitting is performed on the passenger flow density data of each station, including the number of passengers getting on and off and the number of people waiting at the station.

[0056] A convolutional neural network model is used to perform nonlinear fitting on time features, meteorological data, station surrounding event features, passenger boarding and alighting data, and station waiting data to output the predicted passenger flow density value for each station. For example, daily rainfall and wind force level features are extracted from meteorological data, and time period labels are extracted from time data. Time data includes morning and evening peak hours, with the morning peak marked as 1 and the off-peak as 0. Time features also include weekdays or holidays. Station surrounding event features are used to indicate whether there are large-scale gatherings of people around the station. Different values ​​are used to mark the number of people in large-scale gatherings, for example, the specific value of the station surrounding event features is determined according to the scale of a marathon.

[0057] The actual number of passengers at each station is determined based on the number of passengers getting on and off. The station passenger flow sequence is constructed based on the actual number of passengers and the number of people waiting at the station. In addition, the station passenger flow sequence also includes meteorological data, time characteristics, and characteristics of events around the station. A long short-term memory network with a gating mechanism is used to input the station passenger flow sequence of the current and multiple historical time periods and output the passenger flow prediction value for the future time period. The passenger flow prediction value is used to represent the number of passengers who need to take the train at the corresponding station in the future time period.

[0058] The predicted passenger flow density is obtained by dividing the number of passengers by the area of ​​the waiting area at the station. This predicted passenger flow density value represents the congestion level of the corresponding station within the future unit of time. The predicted passenger flow density value is used as the passenger flow density value for each station. For example, when the predicted passenger flow density value of any station exceeds a preset density threshold, a congestion warning signal is generated. The specific value of the preset density threshold can be adjusted according to the actual situation and is not limited here.

[0059] S130. Multiple line section intervals are delineated using road network data, and the upper limit data of line section passenger flow is determined based on the passenger flow density values ​​of each station included in the line section interval.

[0060] The road network data includes information on all stations and routes served by public transportation. The data is divided into multiple route cross-sectional sections, each containing multiple stations. For example, the road network data can be gridded, with each grid corresponding to one route cross-sectional section. The size of the grid can be adjusted according to actual needs and is not limited here; for example, each grid may contain 3-5 stations.

[0061] The passenger flow density value of each station within each route section is statistically analyzed to obtain the total passenger flow density value. This total passenger flow density value is then combined with the length of the route section and the number of stations to obtain the upper limit passenger flow data for that route section. For example, station weights are determined based on the number of different types of public transportation vehicles stopping at each station. The total passenger flow density value for the route section is determined based on the weighted fusion result of the station weights and passenger flow density values. The operating route length and the number of stations for the public transportation vehicles included in the route section are then determined. These operating route lengths, the number of stations, and the total passenger flow density value are input into a pre-trained convolutional neural network model to obtain the upper limit passenger flow data for that route section. This upper limit passenger flow data represents the maximum number of passengers required to be carried within the route section, including both passengers on public transportation vehicles and passengers waiting to board at stations.

[0062] S140. Adjust the scheduling information of public transportation vehicles based on the passenger flow limit data of the line section.

[0063] Since the upper limit data of passenger flow of a line section represents the actual passenger flow pressure of that line section, if the actual passenger flow pressure of a certain line section is too high, the departure interval of the public transportation corresponding to that line section will be reduced; if the actual passenger flow pressure of a certain line section is low, the departure interval of the public transportation corresponding to that line section will be increased, or the operation route of the corresponding public transportation will be adjusted to alleviate the pressure on other line sections with high passenger flow.

[0064] For example, in a practical application scenario, when the upper limit of passenger flow at a line section exceeds a preset passenger flow threshold, a departure interval shortening instruction is triggered to avoid excessive crowding at the section. The preset passenger flow threshold can be adjusted according to actual conditions and is not restricted here. Furthermore, the departure interval is dynamically adjusted based on the constraint that the waiting time at the station is less than a preset waiting threshold, generating a dispatch instruction containing vehicle number, departure time, and temporary detour suggestions.

[0065] In one feasible embodiment, the scheduling information includes departure time information; S140 includes:

[0066] Determine the basic constraints; these constraints are used to determine the minimum and maximum departure intervals for different time periods.

[0067] Based on the passenger flow limit data of the line section and the basic constraints, an objective function is constructed to minimize the total waiting time, thereby obtaining the information on the combination of adjusted departure times for public transportation.

[0068] Set basic constraints, including minimum and maximum departure intervals for different time periods, to ensure the normal operation of public transportation, such as the lower limit of the minimum departure interval during the morning peak hours and the upper limit of the maximum departure interval during the off-peak hours.

[0069] Based on the upper limit data of passenger flow at the line section and basic constraints, a departure time optimization model is constructed. The model aims to minimize the total waiting time at each station from exceeding the preset time threshold. It dynamically adjusts the departure interval and generates dispatch instructions containing vehicle number, departure time and temporary detour suggestions, thus obtaining the adjusted departure time combination information of public transportation vehicles.

[0070] For example, the remaining number of seats, current vehicle speed and fault codes uploaded by the vehicle terminal are received in real time, and dispatchable vehicles are marked. Based on the dispatchable vehicles, according to the upper limit data of passenger flow in the line section and basic constraints, a genetic algorithm is used to solve the optimal departure time combination with the objective function of minimizing the total waiting time, which is used as information for adjusting the departure time combination.

[0071] This embodiment improves resource utilization and passenger satisfaction by using intelligent scheduling based on passenger flow, dynamically adjusting departure intervals based on genetic algorithms and basic constraint settings.

[0072] In one feasible embodiment, the method further includes:

[0073] Structured passenger flow statistics and unstructured video clips are stored in a distributed database; the passenger flow statistics include at least the passenger flow density values ​​of each station and the upper limit data of passenger flow at each line section.

[0074] The scheduling information and passenger flow statistics are displayed in the form of a heat map.

[0075] The passenger flow statistics results are used to statistically analyze the passenger flow-related information obtained in the above steps. This includes at least passenger flow density values ​​for each station and upper limit passenger flow data for each route section. It may also include multi-source data collection results, such as location data, passenger boarding and alighting data, and waiting passenger data for multiple public transportation vehicles. Unstructured video clips refer to video information captured by visual cameras during multi-source data collection. For example, video clips of passengers boarding and alighting captured by visual cameras deployed at vehicle doors, or images of waiting passengers captured by visual cameras deployed at stations. Both structured passenger flow statistics results and unstructured video clips are stored for subsequent traceability and verification. Furthermore, unstructured video clips whose storage time exceeds a preset duration threshold are deleted to improve memory resource utilization. The preset duration threshold can be adjusted according to actual scenario requirements and is not limited here.

[0076] Furthermore, it supports the visualization of public transportation dispatch information and passenger flow statistics through heat maps, which can intuitively reflect the spatiotemporal distribution characteristics of urban traffic operation and passenger travel demand. The heat map uses a map as its base and employs color gradients (such as blue-yellow-red) to represent indicators such as passenger density, number of passengers boarding and alighting, vehicle occupancy rate, or frequency of service in different areas or stations during specific time periods: high-heat areas (such as red) represent dense passenger flow, high-frequency vehicle dispatch, or congestion points, while low-heat areas (such as blue) represent sparse passenger flow or areas with weak service. Simultaneously, dispatch data such as bus route directions, real-time vehicle locations, and departure intervals can be overlaid to achieve dynamic matching analysis of operational efficiency and passenger demand, providing data support for optimizing network layout, adjusting departure plans, and improving the quality of public transportation services.

[0077] For example, HBase is used to store structured passenger flow statistics, while object storage services store unstructured video clips, with data retained for 90 days. Duplicate data is deduplicated from the stored data, and missing values ​​are imputed using the mean of neighboring stations. Data query services are provided via a RESTful API. Data query permissions are assigned according to user roles: dispatchers can access data for the entire line, while site administrators can only access data for their assigned sites.

[0078] This embodiment implements distributed data storage, governance, interface services, and access control to ensure the effective management and secure use of passenger flow data.

[0079] The technical solution of this embodiment improves the accuracy of passenger flow information determination by integrating multi-source data to determine passenger flow information of public transportation vehicles; furthermore, it divides the route section into segments using road network data, analyzes the passenger flow data within the segment, and then schedules public transportation vehicles based on the passenger flow data of the segment, thereby improving the accuracy of scheduling and avoiding errors caused by scheduling based on information from a single station.

[0080] Figure 2 This is a flowchart of another public transportation management method based on multi-source data provided by an embodiment of the present invention. This embodiment further refines the process of determining the upper limit data of passenger flow at the line section in the above embodiments. For example... Figure 2 As shown, the method includes:

[0081] S210 collects real-time location data, passenger boarding and alighting data, station waiting data, and weather data from multiple public transportation vehicles.

[0082] S220. Based on meteorological data, time characteristics, and event characteristics around the station, nonlinear fitting is performed on the passenger flow density data of each station, including the number of passengers getting on and off and the number of people waiting at the station.

[0083] S230 uses road network data to delineate multiple line section intervals.

[0084] S240. Determine the current passenger flow data of the line based on the current passenger flow density values ​​of each station included in the line section.

[0085] The passenger flow density value includes the current passenger flow density value and the historical passenger flow density value. The current passenger flow density value represents the passenger flow density at the current point in time, while the historical passenger flow density value represents the passenger flow density at the same time in the past corresponding to the current point in time. For example, using days as the unit, if the current point in time is 9:00 AM, then the historical passenger flow density value represents the passenger flow density at 9:00 AM the previous day. In addition, the passenger flow density value may also include the predicted passenger flow density value, which represents the passenger flow density at the same time in the future corresponding to the current point in time. For example, the predicted passenger flow density value represents the predicted passenger flow density at 9:00 AM the next day.

[0086] For example, the current passenger flow density value of each station included in each route section is statistically analyzed to obtain the current total passenger flow density value. This is then combined with the length of the route section and the number of stations included to obtain the current route passenger flow data for that section. For instance, station weights are determined based on the number of different types of public transportation vehicles stopping at each station. The current total passenger flow density value for that route section is determined based on the weighted fusion result of the station weights and the current passenger flow density value. The operating route length and the number of stations included in the public transportation vehicles within the route section are also determined. These operating route lengths, the number of stations included, and the current total passenger flow density value are then input into a pre-trained convolutional neural network model to obtain the current route passenger flow data for that route section.

[0087] S250. Determine historical passenger flow data based on the historical passenger flow density values ​​of each station included in the section of the line.

[0088] For example, the historical passenger flow density values ​​of each station included in each route section are statistically analyzed to obtain the historical total passenger flow density value. This is then combined with the length of the route section and the number of stations included to obtain the historical route passenger flow data for that route section. For instance, station weights are determined based on the number of different types of public transportation vehicles stopping at each station. The historical total passenger flow density value for that route section is determined based on the weighted fusion result of the station weights and historical passenger flow density values. The operating route lengths and the number of stations included in the public transportation vehicles within the route section are then determined. The operating route lengths, the number of stations included, and the historical total passenger flow density value are then input into a pre-trained convolutional neural network model to obtain the historical route passenger flow data for that route section.

[0089] Optionally, multiple historical passenger flow density values ​​for each station within each route section can be statistically analyzed to obtain multiple historical total passenger flow density values. These values, combined with the length of the route section and the number of stations included, yield multiple historical route passenger flow data for that section. The average of these historical passenger flow data is then used as the final historical route passenger flow data. For example, historical route passenger flow data for the same period over the past 30 days can be obtained, and a reference baseline value can be generated using exponential smoothing to serve as the final historical route passenger flow data.

[0090] S260. Determine the current weight of the current route passenger flow data based on the difference between the current route passenger flow data and the historical route passenger flow data, and determine the historical weight of the historical route passenger flow data based on the current weight.

[0091] Since current passenger flow data reflects the number of passengers required to be carried in the current section of the line, while historical passenger flow data reflects the number of passengers required to be carried in the same section of the line during the same period in the past, if there is a significant increase in passenger flow compared to the same period in the past, it means that the upper limit of the number of passengers required to be carried in the current section of the line needs to be adjusted according to the specific circumstances of the current time.

[0092] For example, the difference between the current route passenger flow data and the historical route passenger flow data is determined. If the difference is greater than a preset passenger flow increment threshold, the current weight is assigned a first value, and the historical weight is assigned a second value, where the first value is greater than the second value. If the difference is less than or equal to the preset passenger flow increment threshold, the current weight is assigned a third value, and the historical weight is assigned a fourth value, where the first value is greater than the third value. The relative values ​​of the third and fourth values ​​are not restricted, and the sum of the current and historical weights is 1. For example, the first value is 0.9, the second value is 0.1, the third value is 0.6, and the fourth value is 0.4.

[0093] S270. Determine the upper limit data of passenger flow at the line section based on the current line passenger flow data and current weight, as well as the weighted result of historical line passenger flow data and historical weight.

[0094] The upper limit of passenger flow at the line section is determined using a weighted algorithm that integrates historical data from the same period. Specifically, this includes: The current weight Adaptive coefficients When a sudden increase in passenger flow at a site exceeds a preset threshold, the current weight is automatically increased to 0.9. When a sudden event is detected at a site, [the weight will be adjusted accordingly]. The value increases linearly from the default 0.7 to 0.9, and then returns to the default value after a preset duration following the event. Unexpected events include the end of a concert or other events with high traffic.

[0095] The upper limit data for passenger flow at a specific section of the line reflects the actual passenger flow pressure at that section. With a waiting time at stations not exceeding 5 minutes as a constraint, the departure interval is dynamically adjusted based on the magnitude of the upper limit data. In practical applications, when the upper limit data exceeds a preset threshold, a command to shorten the departure interval is triggered to prevent overcrowding at that section. Furthermore, the calculation of the upper limit data integrates current real-time passenger flow with historical data from the same period, reflecting both the dynamic changes in current passenger flow and historical patterns. This data can be used to analyze the periodicity of passenger flow on the line. Sudden characteristics, such as comparing the passenger flow limit data of line sections at different time periods, can identify peak passenger flow periods and high-incidence sections, providing data support for line planning and capacity allocation. When the sudden increase in passenger flow at a station exceeds the preset passenger flow increase threshold, the current weight is automatically increased to the upper limit value, making the passenger flow limit data of the line section more focused on the current real-time data. At this time, the passenger flow limit data of the line section can quickly reflect the passenger flow pressure of the section under the sudden situation, prompting the system to prioritize the generation of scheduling strategies based on the current emergency, such as increasing temporary train services and adjusting detour routes, thereby improving the response efficiency to emergencies.

[0096] Furthermore, the predicted passenger flow data of the line section is determined based on the predicted passenger flow density values ​​of each station included in the line section. The upper limit data of the line section passenger flow is determined based on the weighted result of the current line passenger flow data and current weight, historical line passenger flow data and historical weight, and predicted line passenger flow data and predicted weight. The value of the predicted weight is less than the current weight and historical weight, and the sum of the predicted weight, current weight, and historical weight is 1.

[0097] S280. Adjust the dispatching information of public transportation vehicles based on the upper limit data of passenger flow at the line section.

[0098] The technical solution of this embodiment determines the upper limit data of passenger flow at the line section by integrating current line passenger flow data and historical line passenger flow data. This not only reflects the dynamic changes of current passenger flow but also takes into account historical patterns, which facilitates the improvement of the accuracy of analyzing the periodicity of line passenger flow. Furthermore, the integration weight is determined based on the difference between current line passenger flow data and historical line passenger flow data during the integration process, which facilitates the improvement of the accuracy of analyzing the sudden characteristics of line passenger flow.

[0099] Figure 3 This is a flowchart of another public transportation management method based on multi-source data provided by an embodiment of the present invention. This embodiment further refines the process of determining passenger flow density values ​​in the above embodiments. Figure 3 As shown, the method includes:

[0100] S310 collects real-time location data, passenger boarding and alighting data, station waiting data, and weather data from multiple public transportation vehicles.

[0101] S320. Based on the current and historical passenger boarding and alighting data and station waiting data, the first neural network model determines the passenger flow forecast value.

[0102] The system uses a first neural network model as input to the current and historical station passenger flow sequences and outputs predicted passenger flow values ​​for future times. The first neural network model can be a long short-term memory network with a gating mechanism or other neural network models. The station passenger flow sequence includes at least the number of people waiting at the station and the number of people getting on and off. For example, the station passenger flow sequence is a numerical vector in which the first value represents the number of people waiting at the station, the second value represents the number of people getting on, and the third value represents the number of people getting off.

[0103] For example, the formula for predicting passenger flow is: ;in, This is a forecast of station passenger flow for the next hour. This is a historical passenger flow sequence for the station, spanning 3 hours from the current location. The time period weight coefficients are used to improve the output of the LSTM and enhance the model's sensitivity to periodic peaks. (For example, the first value is used for the morning peak, the second value for the off-peak, and the third value for the evening peak, where the first value is greater than the second value and the third value is greater than the second value.) The event-specific passenger flow correction value is determined based on the characteristics of events surrounding the site. For example, when a large-scale event is detected near the site, a different positive value is dynamically assigned to the event's passenger flow correction value based on the event's scale (such as the number of participants). (For example, for every additional 1000 participants, the correction value is adjusted accordingly.) (Add 50-200 people).

[0104] Optionally, when the time point corresponding to the passenger flow forecast arrives, the actual passenger flow value at that time point is obtained, and the weight parameters of the first neural network model are adjusted based on the error between the passenger flow forecast and the actual passenger flow value. For example, when the error between the passenger flow forecast and the actual passenger flow value exceeds a preset error threshold, the historical data retraining mechanism of the first neural network model is automatically triggered to adjust the model weight parameters.

[0105] For example, the error correction formula is as follows: ;in, These are the updated weight parameters of the first neural network model. These are the current weight parameters of the first neural network model. Set the learning rate (for example, a value of 0.01). The error gradient is calculated based on the deviation between the predicted passenger flow and the actual passenger flow. When , the value is 1; the value of 'a' can be adjusted according to the actual situation and is not restricted here.

[0106] S330. Determine the real-time passenger flow value based on the current passenger boarding and alighting data and the number of people waiting at the station.

[0107] The real-time passenger flow value is determined by a third neural network model based on the current passenger boarding / alighting data and the station waiting data. For example, the third neural network model takes the current station passenger flow sequence as input and outputs the current real-time passenger flow value. The third neural network model can be a convolutional neural network. The station passenger flow sequence includes at least the station waiting data and the passenger boarding / alighting data. For instance, the station passenger flow sequence is a numerical vector where the first value represents the station waiting data, the second value represents the passenger boarding data, and the third value represents the passenger alighting data.

[0108] Optionally, to improve calculation efficiency, the real-time passenger flow value can be determined based on the difference between the number of passengers getting on and off, and the sum of the number of people waiting at some stations. The specific proportion of the number of people waiting at some stations can be determined based on the number of public transportation vehicles stopping at that station.

[0109] S340. Determine the predicted passenger flow density value based on the passenger flow forecast value.

[0110] The predicted passenger flow density is determined by the ratio of the predicted passenger flow to the waiting area area of ​​the station.

[0111] S350. Determine the real-time passenger flow density value based on the real-time passenger flow value.

[0112] The real-time passenger flow density is determined by the ratio of the real-time passenger flow value to the waiting area area of ​​the station.

[0113] In one feasible embodiment, determining the real-time passenger flow density value based on the real-time passenger flow value includes:

[0114] The correction coefficient is determined by the second neural network model based on time characteristics, meteorological data, and event characteristics around the station; the real-time passenger flow density value is determined by multiplying the ratio of the real-time passenger flow value to the waiting area area of ​​the station with the correction coefficient.

[0115] The formula for calculating the real-time passenger flow density is as follows: ,in, This is the real-time passenger flow density value. This refers to the real-time passenger flow value determined by data on the number of passengers getting on and off and the number of people waiting at stations. The area of ​​the waiting area at the station. To determine the correction coefficients based on the output of the second neural network model, the input parameters of the second neural network model are: The time feature vector represents the current moment, including whether it is a weekday and whether it is the morning rush hour (e.g., weekday is marked 1, holiday is marked 0, morning rush hour is marked 1, off-peak is marked 0; this value is only an example and different values ​​can be used to represent it). This represents the meteorological feature vector at the current moment (e.g., the meteorological feature vector includes daily rainfall and wind force level). The second neural network model can be a convolutional neural network model or any improved convolutional neural network model. It represents the characteristics of events around the site at the current moment (e.g., the time characteristics around the site include the quantitative value of the scale of large events).

[0116] Furthermore, the predicted passenger flow density value is determined based on the passenger flow forecast value, including:

[0117] The second correction coefficient is determined by the fourth neural network model based on time characteristics, meteorological data, and event characteristics around the station; the predicted passenger flow density value is determined by multiplying the ratio of the predicted passenger flow value to the waiting area area of ​​the station with the second correction coefficient.

[0118] The specific process can be referred to above for correcting the real-time passenger flow density value based on the correction coefficient, and will not be repeated here.

[0119] S360: Multiple line section intervals are delineated through road network data, and the upper limit data of line section passenger flow is determined based on the passenger flow density value of each station included in the line section interval.

[0120] S370: Adjust the dispatching information of public transportation vehicles based on the upper limit data of passenger flow at the line section.

[0121] The technical solution of this embodiment determines the passenger flow prediction value at historical times and the passenger flow prediction value at the current time, and then determines the real-time passenger flow density value and the predicted passenger flow density value. It improves the comprehensiveness of the passenger flow density value from both current and historical perspectives, which facilitates the improvement of the accuracy of analyzing the periodicity of passenger flow on the route.

[0122] Figure 4 The diagram shown is an architectural schematic of a public transportation vehicle management system based on multi-source data provided by an embodiment of the present invention. Figure 4 As shown, the system includes:

[0123] The system includes a multi-source data access module, a dynamic passenger flow processing module, a cross-sectional analysis module, an intelligent scheduling module, a data platform core module, and a terminal display module. These modules are connected through network communication and interfaces to achieve data transmission and interaction.

[0124] The multi-source data access module includes a data acquisition module, which is used to collect real-time data from vehicle GPS positioning, passenger flow sensors (passenger boarding and alighting data), station video surveillance (waiting passenger data), and meteorological data. The multi-source data access module transmits the raw data to the data platform in a streaming manner through network communication.

[0125] In addition, the multi-source data access module also includes a trajectory cleaning unit, which is used to filter the positioning data using Kalman filtering, determine the current trajectory information based on the filtered positioning data, and correct the current trajectory information based on the matching result of historical trajectory information and current trajectory information.

[0126] The passenger flow sensor verification unit is used to filter the passenger boarding and alighting data based on the timestamp synchronization results of the positioning data and the passenger boarding and alighting data.

[0127] The video stream decoding unit receives and parses the station monitoring video, and through face detection preprocessing, only transmits key frames containing waiting crowds to the central platform for statistical analysis of the number of people waiting at the station.

[0128] The data standardization unit is connected to the three units mentioned above to convert data in different formats into a JSON structure. The passenger flow data field includes the station ID, number of passengers boarding, number of passengers alighting, number of people waiting at the station, and the collection timestamp.

[0129] like Figure 5 The diagram shown is a structural schematic of a multi-source data access module. Figure 6 The diagram shows the data processing procedure of the multi-source data access module.

[0130] The dynamic passenger flow processing module is used to perform nonlinear fitting on the passenger boarding and alighting data and the waiting passenger data at each station based on the meteorological data, time characteristics, and event characteristics around the station, so as to obtain the passenger flow density value of each station.

[0131] The dynamic passenger flow processing module specifically includes a feature extraction unit, which is used to extract relevant feature information from meteorological data and time data.

[0132] The time-series establishment unit is used to determine the passenger flow prediction value based on the passenger boarding and alighting data and the station waiting data at the current time and historical time using the first neural network model.

[0133] The error correction unit is used to connect with the time series establishment unit. When the error between the predicted value and the actual value exceeds the error threshold, the historical data retraining mechanism is automatically triggered to adjust the weight parameters of the first neural network model.

[0134] The density calculation unit is used to determine the predicted passenger flow density value based on the ratio of the predicted passenger flow value to the waiting area area of ​​the station; and to determine the real-time passenger flow density value based on the ratio of the real-time passenger flow value to the waiting area area of ​​the station.

[0135] The dynamic passenger flow processing module is designed for the operation scenarios of public transportation vehicles. By integrating multi-source features, strengthening periodic and event responses, and dynamically correcting errors, it improves the traditional CNN and LSTM algorithms, thereby enhancing the accuracy of passenger flow analysis and prediction.

[0136] The cross-section analysis module is used to delineate multiple route cross-section intervals using road network data, and determine the upper limit data of the route cross-section passenger flow based on the passenger flow density values ​​of each station included in the route cross-section interval. The cross-section analysis module is connected to the dynamic passenger flow processing module through a microservice interface, and calculates the upper limit data of the route cross-section passenger flow based on the station passenger flow density values, using a weighted algorithm that integrates historical data from the same period.

[0137] The cross-section analysis module specifically includes a cross-section identification unit, which is used to delineate the cross-section intervals of the line using GIS road network data.

[0138] The traffic aggregation unit, connected to the cross-section identification unit, accumulates the number of passengers getting on and off at each station within the cross-section and calculates the net traffic difference.

[0139] The historical data retrieval unit is used to obtain historical passenger flow data for the same period over the past 30 days, and generates a reference baseline value using exponential smoothing as the final historical passenger flow data.

[0140] The adaptive coefficient adjustment unit is used to adjust the passenger flow limit data of the line section when a sudden event is detected at the station. The value increases linearly from the default 0.7 to 0.9, and then returns to the default value after a preset duration following the event. Unexpected events include the end of a concert or other events with high traffic.

[0141] The intelligent scheduling module is used to adjust the scheduling information of the public transportation vehicles based on the passenger flow limit data of the route section. Specifically, it receives the output results of the section analysis module, constructs a departure time optimization model, which dynamically adjusts the departure interval with the waiting time at the station as a constraint, and generates scheduling instructions containing vehicle number, departure time, and temporary detour suggestions.

[0142] The intelligent dispatch module also includes a vehicle status monitoring unit, which receives the remaining number of seats, current vehicle speed and fault codes uploaded by the on-board terminal in real time, and marks vehicles that can be dispatched.

[0143] The constraint setting unit allows you to set basic constraints, such as a minimum departure interval greater than 2 minutes during the morning peak hours and a maximum interval less than 10 minutes during off-peak hours.

[0144] The genetic algorithm solution unit is connected to the vehicle status monitoring unit and the constraint setting unit, respectively. It takes minimizing the total waiting time as the objective function and outputs the optimal combination of departure times.

[0145] The instruction generation unit converts the solution results into structured instructions containing vehicle ID, departure time, and stops along the route, and pushes them to the driver's terminal via the Internet.

[0146] The core module of the data platform is used to store structured passenger flow statistics and unstructured video clips through a distributed database. The passenger flow statistics include at least the passenger flow density values ​​for each station and the upper limit of passenger flow for each route section. For example, it interacts with the aforementioned module through a distributed database to store structured passenger flow statistics and unstructured video clips, providing data query and interface services.

[0147] The core modules of the data platform also include: a distributed storage unit, which uses a database to store structured passenger flow statistics and an object storage service to store video clips, with a data retention period of 90 days.

[0148] The data governance unit performs deduplication on duplicate data collection and uses the mean of neighboring sites to impute missing values.

[0149] The Interface Service Unit provides an information exchange interface and supports querying passenger flow data by site ID and time period.

[0150] Access Control Unit: Connects to the Interface Service Unit to assign data access permissions to different user roles. Dispatchers can view data for the entire line, while site administrators can only view data for the sites under their jurisdiction.

[0151] like Figure 7 The diagram shown is a structural schematic of the core module of the data platform.

[0152] The terminal display module is used to display the scheduling information and passenger flow statistics in the form of a heat map. For example, upon receiving instructions from the intelligent scheduling module, it can visualize the distribution of passenger flow at stations and the real-time location of vehicles in the form of a heat map, allowing relevant personnel to intuitively understand the traffic operation status.

[0153] like Figure 8 The diagram shows the data processing procedures of the dynamic passenger flow processing module, cross-sectional analysis module, and intelligent scheduling module. Figure 9 The diagram shown illustrates the data processing procedure of the core module of the data platform. Figure 10The diagram shows the runtime sequence of a public transportation management system based on multi-source data. Specifically, it includes: First, deploying edge computing nodes at the data acquisition layer to perform real-time cleaning and standardization (using JSON format) of heterogeneous data such as GPS, passenger flow, and video. The pre-processed data is then uploaded to the data platform via a 5G network. Second, the dynamic passenger flow processing module uses an LSTM-TCN hybrid model for temporal feature extraction, combined with a GAN (Generative Adversarial Network) to enhance training data, achieving minute-level passenger flow density prediction and anomaly detection. Next, the cross-sectional analysis module calls a computer vision API to parse surveillance video, uses a graph convolutional network (GCN) to construct a station association graph, and outputs passenger flow prediction results including the impact of sudden events. The intelligent scheduling module generates multi-objective optimization schemes based on reinforcement learning algorithms, and after simulation verification, sends scheduling instructions to the vehicle terminal via a microservice API. Finally, the data platform updates data assets through a stream-batch integrated processing engine, and the large-screen visualization system renders operational indicators in real time, forming a closed-loop iteration of "data acquisition - model training - decision optimization - effect evaluation." The process needs to be implemented in three phases: first, complete the infrastructure upgrade and data governance; second, deploy the core algorithm and conduct small-scale verification; and finally, achieve system-wide joint debugging and performance optimization.

[0154] The system in this embodiment includes: a multi-source data fusion processing mechanism, which effectively collects, cleans, verifies, and standardizes data from vehicle GPS, passenger flow sensors, station video surveillance, and meteorological data through a multi-source data access module, providing a comprehensive and accurate data foundation for subsequent analysis and scheduling; a dynamic passenger flow analysis and prediction method based on an improved algorithm, combined with a convolutional neural network, considering multiple factors such as time and weather, to calculate real-time passenger flow density at stations and predict future passenger flow, with error correction capabilities; a cross-sectional passenger flow calculation method that integrates historical data, using a weighted algorithm to calculate the maximum cross-sectional passenger flow of the route, and improving the accuracy and adaptability of cross-sectional passenger flow calculation through adaptive coefficient adjustment; intelligent scheduling based on passenger flow conditions, dynamically adjusting departure intervals based on genetic algorithms and constraint settings, generating optimized scheduling instructions, improving resource utilization and passenger satisfaction; and a high-efficiency data platform core module architecture, realizing distributed data storage, governance, interface services, and access control, ensuring effective data management and secure use.

[0155] Therefore, the system in this embodiment can achieve the following: By employing a multi-source data access module, it integrates various data sources such as vehicle GPS, passenger flow sensors, station video surveillance, and meteorological data, providing a more comprehensive reflection of passenger flow and the operating environment, thus improving data accuracy. Through various units and algorithms in the dynamic passenger flow processing module, it can perform in-depth analysis and prediction of passenger flow at each station, considering factors such as time and weather, making passenger flow analysis and prediction more accurate and providing strong data support for operation management. Based on real-time and cross-sectional passenger flow, a genetic algorithm is used to solve for the optimal departure time, dynamically adjusting the departure interval. This minimizes the total waiting time while meeting station waiting time constraints, improving resource utilization and reducing congestion during peak hours and capacity waste during off-peak hours. The core module of the data platform implements distributed data storage, governance, interface services, and access control, improving data processing and reuse efficiency, while ensuring secure data use through access control.

[0156] The solution of this invention can not only be applied to bus operations, but its logic of multi-source data fusion, dynamic passenger flow analysis, and intelligent scheduling can also be transferred to other public transportation fields such as subways, light rail, and BRT (Bus Rapid Transit). For example, the subway system can reuse the "multi-source data access module" to integrate gate passenger flow data, platform video surveillance data, train GPS data, and meteorological data. The "cross-section analysis module" calculates the passenger flow pressure in the platform and tunnel sections, and the "intelligent scheduling module" dynamically adjusts the train departure interval or temporarily adds extra trains to solve the congestion problem during morning and evening rush hours.

[0157] Furthermore, the core module of the data platform of this invention can be integrated with the urban traffic command system to achieve coordinated management and control of public transportation, private vehicles, shared bicycles, and other modes of transportation. For example, when passenger flow at bus stops surges, the system can link with nearby shared bicycle dispatch platforms to temporarily deploy additional vehicles to supplement capacity; or push detour suggestions to private vehicles to reduce the impact of traffic congestion around bus stops on bus punctuality.

[0158] The solution in this embodiment integrates GPS, passenger flow sensors, video surveillance and meteorological data, and can also incorporate user travel APP data (such as bus APP ticket booking and arrival reminder records) and city event data (such as large-scale exhibitions and sports event schedules) to further improve the timeliness and accuracy of passenger flow forecasting.

[0159] The dynamic passenger flow processing module uses an improved convolutional neural network combined with a long short-term memory network model. It can also use reinforcement learning algorithms to continuously optimize the prediction model through real-time feedback from actual scheduling effects (such as passenger waiting time and vehicle occupancy rate), thereby improving the adaptive capability in complex scenarios.

[0160] In this embodiment of the invention, a privacy desensitization unit can be added to the core module of the data platform to blur the facial information of the video surveillance (retaining the outline features for passenger flow statistics, but not storing details that can identify individuals); and to perform spatiotemporal anonymization processing on the GPS trajectory data (such as removing the direct association between specific vehicle numbers and timestamps, and only retaining trajectory aggregation data at the "route-time period" level).

[0161] The specific operation process of this system includes:

[0162] The first step involves real-time collection of location data from the bus's GPS, passenger flow data from onboard sensors, waiting passenger data from station video surveillance, and weather data. This raw data is then transmitted to the data platform via a 5G message queue protocol using streaming processing. The second step involves dynamic passenger flow processing of the standardized data: an improved convolutional neural network model is used to fuse time features, weather factors, and station-related event characteristics for nonlinear fitting, outputting real-time passenger flow density values ​​for each station. The third step calculates the upper limit of passenger flow for the route section based on the passenger flow density values ​​at each station: a weighted algorithm incorporating historical data from the same period is employed. When a sudden increase in passenger flow at a station exceeds 30%, the current weight is automatically increased to 0.9; the fourth step is to build a departure time optimization model: dynamically adjust the departure interval with the constraint that the waiting time at the station is less than 5 minutes, and generate dispatch instructions containing vehicle number, departure time and temporary detour suggestions; the fifth step is to store structured passenger flow statistics and unstructured video clips through a distributed database, and provide data query and interface services; the sixth step is to visualize the dispatch instructions and passenger flow data in the form of a heat map through the WebSocket protocol.

[0163] The first step specifically includes: filtering drift points in the vehicle GPS positioning data, retaining the trajectory within the 0-60km / h speed range, and correcting positioning errors using Kalman filtering; matching the door opening / closing signals with the station location timestamps to remove erroneous passenger data caused by vehicles not stopping properly; parsing the station monitoring video using H.265 encoding, performing face detection preprocessing at edge nodes, and transmitting only keyframes containing waiting passengers; and converting multi-source data into JSON format, with passenger flow data fields including station ID, number of passengers boarding / alighting, and collection timestamp.

[0164] The second step specifically includes: extracting daily rainfall and wind force characteristics from meteorological data, and extracting time period labels from time data; using an LSTM network with a gating mechanism, inputting the passenger flow sequence of the previous 3 hours, and outputting the passenger flow prediction value for the next 1 hour; when the error between the predicted value and the actual value exceeds 15%, triggering the historical data retraining mechanism to adjust the model weights; calculating the ratio of the number of passengers to the waiting area area to obtain the passenger flow density value, and generating a congestion warning when it exceeds 0.8 people / m².

[0165] The third step specifically includes: delineating a section of the line containing 3 to 5 consecutive stations based on GIS road network data; calculating the net flow difference by accumulating the number of passengers getting on and off at each station within the section; calling up passenger flow data from the same section over the past 30 days and generating a reference benchmark value through exponential smoothing; when a sudden event is detected at a station, linearly increasing the current weight value from 0.7 to 0.9, and restoring it 15 minutes after the event ends.

[0166] The fourth step specifically includes: real-time acquisition of the remaining number of seats, vehicle speed, and fault codes uploaded by the vehicle terminal, and marking vehicles that can be dispatched; setting basic constraints: the minimum departure interval during the morning peak is greater than 2 minutes, and the maximum interval during the off-peak period is less than 10 minutes; using a genetic algorithm to solve for the optimal combination of departure times with the goal of minimizing the total waiting time; and converting the solution into a structured instruction containing vehicle ID, departure time, and stops along the route and pushing it to the driver's terminal.

[0167] The fifth step specifically includes: using HBase to store structured passenger flow data, and object storage services to save video clips, with data retention for 90 days; performing deduplication on duplicate data and using the mean of neighboring sites to fill in missing values; providing data query services with an interface response time of less than 500ms; and assigning data permissions according to user roles, with dispatchers able to access data for the entire line and site administrators only able to access data for their managed sites.

[0168] This invention provides a research on a method for intelligent fusion and unified management of multi-source data in discrete systems based on data governance. It addresses the problem in existing technologies where data collection sources are singular, relying solely on vehicle-mounted passenger flow sensors and neglecting other key data such as station video surveillance and weather data, resulting in incomplete and inaccurate judgments of passenger flow. This invention achieves a comprehensive reflection of passenger flow and operating environment conditions, thereby improving data accuracy.

[0169] Figure 11 This is a schematic diagram of a public transportation vehicle management device based on multi-source data, provided as an embodiment of the present invention. Figure 11 As shown, the device includes:

[0170] The multi-source data acquisition module 1101 is used to collect real-time location data, passenger boarding and alighting data, station waiting data, and meteorological data from multiple public transportation vehicles.

[0171] The station passenger flow density value determination module 1102 is used to perform nonlinear fitting on the passenger boarding and alighting data and the station waiting number data based on the meteorological data, time characteristics, and station surrounding event characteristics to obtain the passenger flow density value of each station.

[0172] The line section passenger flow upper limit data determination module 1103 is used to delineate multiple line section intervals through road network data, and determine the line section passenger flow upper limit data based on the passenger flow density value of each station included in the line section interval.

[0173] The scheduling adjustment module 1104 is used to adjust the scheduling information of the public transportation vehicles based on the passenger flow limit data of the line section.

[0174] The technical solution of this embodiment improves the accuracy of passenger flow information determination by integrating multi-source data to determine passenger flow information of public transportation vehicles; furthermore, it divides the route section into segments using road network data, analyzes the passenger flow data within the segment, and then schedules public transportation vehicles based on the passenger flow data of the segment, thereby improving the accuracy of scheduling and avoiding errors caused by scheduling based on information from a single station.

[0175] Optional, the module for determining the upper limit of passenger flow for line sections is specifically used for:

[0176] The current passenger flow data of the line is determined based on the current passenger flow density value of each station included in the line section.

[0177] Historical passenger flow data is determined based on the historical passenger flow density values ​​of each station included in the cross-sectional section of the line;

[0178] The current weight of the current route passenger flow data is determined based on the difference between the current route passenger flow data and the historical route passenger flow data, and the historical weight of the historical route passenger flow data is determined based on the current weight.

[0179] The upper limit data of passenger flow at the line section is determined based on the current line passenger flow data and the current weight, as well as the weighted result of the historical line passenger flow data and the historical weight.

[0180] Optional, the station passenger flow density value determination module includes:

[0181] The passenger flow prediction unit is used to determine the passenger flow prediction value based on the current and historical passenger boarding and alighting data and the station waiting passenger data using a first neural network model.

[0182] The real-time passenger flow determination unit is used to determine the real-time passenger flow value based on the current passenger boarding and alighting data and the station waiting number data.

[0183] A predicted passenger flow density value determination unit is used to determine a predicted passenger flow density value based on the predicted passenger flow value;

[0184] The real-time passenger flow density value determination unit is used to determine the real-time passenger flow density value based on the real-time passenger flow value.

[0185] Optional, the real-time passenger flow density determination unit includes:

[0186] The correction coefficients are determined by a second neural network model based on the time characteristics, the meteorological data, and the characteristics of events surrounding the station.

[0187] The real-time passenger flow density value is determined by multiplying the ratio of the real-time passenger flow value to the waiting area area of ​​the station with the correction coefficient.

[0188] Optionally, the device also includes a multi-source data processing module, used to perform nonlinear fitting on the passenger boarding / alighting data and station waiting area data based on the meteorological data, time characteristics, and station surrounding event characteristics.

[0189] The positioning data is filtered using a Kalman filter, and the current trajectory information is determined based on the filtered positioning data.

[0190] The current trajectory information is corrected based on the matching results between historical trajectory information and current trajectory information;

[0191] The passenger boarding and alighting data is filtered based on the timestamp synchronization results of the location data and the passenger boarding and alighting data.

[0192] Optionally, the scheduling information includes departure time information; the scheduling adjustment module is specifically used for:

[0193] Determine the basic constraints; these constraints are used to determine the minimum and maximum departure intervals for different time periods.

[0194] Based on the passenger flow limit data of the aforementioned line section and the aforementioned basic constraints, an objective function is constructed to minimize the total waiting time, thereby obtaining the information on the adjusted departure time combinations of public transportation vehicles.

[0195] Optionally, the device may further include a data display module:

[0196] Structured passenger flow statistics and unstructured video clips are stored in a distributed database; wherein the passenger flow statistics include at least the passenger flow density values ​​of each station and the upper limit data of the passenger flow of the line section.

[0197] The scheduling information and passenger flow statistics are displayed in the form of a heat map.

[0198] The public transportation vehicle management device based on multi-source data provided in the embodiments of the present invention can execute the public transportation vehicle management method based on multi-source data provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the method.

[0199] The acquisition, storage, use, and processing of data in this application comply with relevant national laws and regulations and do not violate public order and good morals.

[0200] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0201] Figure 12 A schematic diagram of an electronic device 10, which can be used to implement embodiments of the present invention, is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0202] like Figure 12 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0203] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0204] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods described above, such as public transportation management methods based on multi-source data.

[0205] In some embodiments, the public transportation management method based on multi-source data can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the public transportation management method based on multi-source data described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform the public transportation management method based on multi-source data by any other suitable means (e.g., by means of firmware).

[0206] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific reference products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0207] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0208] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0209] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0210] The systems and technologies described herein can be implemented in computing systems that include back-end components (e.g., as data servers), or computing systems that include switching components (e.g., application servers), or computing systems that include front-end components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such back-end, switching, or front-end components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0211] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0212] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication unit 19, or installed from storage unit 18, or installed from ROM 12. When the computer program is executed by processor 11, it performs the functions defined in the methods of the embodiments of the present invention.

[0213] This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the public transportation management method based on multi-source data as provided in any embodiment of this application.

[0214] In implementing the computer program product, computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0215] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0216] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A public transportation management method based on multi-source data, characterized in that, The method comprises: Real-time collection of positioning data, passenger number data, station waiting number data and meteorological data of multiple public transport vehicles; Nonlinear fitting of the passenger number data and the station waiting number data according to the meteorological data, time characteristics and station surrounding event characteristics to obtain passenger flow density values of each station; Division of multiple line section intervals through road network data, and determination of line section passenger flow upper limit data according to the passenger flow density values of each station included in the line section intervals; Adjustment of scheduling information of the public transport vehicles according to the line section passenger flow upper limit data.

2. The method of claim 1, wherein, The determination of the line section passenger flow upper limit data according to the passenger flow density values of each station included in the line section intervals comprises: Determination of current line passenger flow data according to current passenger flow density values of each station included in the line section intervals; Determination of historical line passenger flow data according to historical passenger flow density values of each station included in the line section intervals; Determination of a current weight of the current line passenger flow data according to a difference between the current line passenger flow data and the historical line passenger flow data, and determination of a historical weight of the historical line passenger flow data according to the current weight; Determination of the line section passenger flow upper limit data according to a weighted result of the current line passenger flow data and the current weight, and the historical line passenger flow data and the historical weight.

3. The method of claim 1, wherein, The nonlinear fitting of the passenger number data and the station waiting number data according to the meteorological data, time characteristics and station surrounding event characteristics to obtain passenger flow density values of each station comprises: Determination of passenger flow prediction values through a first neural network model according to the passenger number data and the station waiting number data at a current time and historical times; Determination of passenger flow real-time values according to the passenger number data and the station waiting number data at the current time; Determination of prediction passenger flow density values according to the passenger flow prediction values; Determination of real-time passenger flow density values according to the passenger flow real-time values.

4. The method of claim 3, wherein, The determination of the real-time passenger flow density values according to the passenger flow real-time values comprises: Determination of a correction coefficient through a second neural network model according to the time characteristics, the meteorological data and the station surrounding event characteristics; Determination of the real-time passenger flow density values according to a product of a ratio of the passenger flow real-time values to a station waiting area and the correction coefficient.

5. The method of claim 1, wherein, Before the nonlinear fitting of the passenger number data and the station waiting number data according to the meteorological data, time characteristics and station surrounding event characteristics, the method further comprises: Filtering of the positioning data using Kalman filtering to determine current trajectory information according to the filtered positioning data; Correction of the current trajectory information according to a matching result of historical trajectory information and the current trajectory information; Screening of the passenger number data according to a timestamp synchronization result of the positioning data and the passenger number data.

6. The method of claim 1, wherein, The scheduling information comprises departure time information; The adjustment of the scheduling information of the public transport vehicles according to the line section passenger flow upper limit data comprises: determine a basic constraint condition, wherein the basic constraint condition is used to determine minimum headway and maximum headway of different time periods; construct a target function according to the line section passenger flow upper limit data and the basic constraint condition, and obtain the adjusted departure time combination information of the public transport vehicle with the minimum total waiting time as the target.

7. The method of claim 1, wherein, The method further comprises: store the structured passenger flow statistics and the unstructured video clips through a distributed database; wherein the passenger flow statistics at least include passenger flow density values of each station and the line section passenger flow upper limit data; display the scheduling information and the passenger flow statistics in the form of a heat map.

8. A public transportation management device based on multi-source data, characterized by, The device comprises: a multi-source data acquisition module, configured to acquire positioning data, boarding and alighting passenger number data, station waiting passenger number data, and meteorological data of a plurality of public transport vehicles in real time; a station passenger flow density value determination module, configured to perform nonlinear fitting on the boarding and alighting passenger number data and the station waiting passenger number data according to the meteorological data, time characteristics, and station surrounding event characteristics, and obtain passenger flow density values of each station; a line section passenger flow upper limit data determination module, configured to demarcate a plurality of line section intervals through road network data, and determine line section passenger flow upper limit data according to the passenger flow density values of each station included in the line section intervals; a scheduling adjustment module, configured to adjust scheduling information of the public transport vehicle according to the line section passenger flow upper limit data.

9. An electronic device, comprising: The electronic device comprises: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the multi-source data based public transport vehicle management method in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions, and the computer instructions are used to enable the processor to implement the multi-source data based public transport vehicle management method in any one of claims 1-7 when executed.

Images