Rail transit shift interval optimization system based on passenger real-time data analysis

Abstract

The invention discloses a rail transit run interval optimization system based on passenger real-time data analysis, and relates to the technical field of rail transit, and the system comprises a data obtaining module which obtains the number of waiting people in a waiting area; the trend prediction module performs analysis to obtain a train waiting prediction function for predicting the number of passengers waiting for the train in the future; the data analysis module is used for predicting to obtain the estimated number of people waiting for the bus in the bus waiting area after the preset time; a station classification module; the optimization simulation module is used for obtaining an estimated optimized number of people; and the scheme screening module is used for carrying out shift interval optimization by using the target time interval scheme. A data acquisition module, a trend prediction module, a data analysis module, an optimization simulation module and a scheme screening module are arranged, so that the rail transit shift interval is adjusted as small as possible while the passenger flow requirement is met.

Description

Technical Field

[0001] This invention relates to the field of rail transit technology, specifically to a rail transit interval optimization system based on real-time passenger data analysis. Background Technology

[0002] Rail transit refers to a type of transportation or system where vehicles operate on specific tracks. Rail transit is generally divided into three main categories: railway systems, intercity rail transit, and urban rail transit. Rail transit generally boasts advantages such as large capacity, high speed, frequent service, safety and comfort, high punctuality, all-weather operation, low costs, and energy efficiency and environmental friendliness. Urban rail transit is essentially the subway. Because passenger flow varies over time, operating rail transit at predetermined time intervals can easily lead to overcrowding at some stations. However, current technology is insufficient for optimizing rail transit service intervals, making it difficult to develop a reasonable automated adjustment scheme. Summary of the Invention

[0003] To address the aforementioned technical problems, this technical solution provides a rail transit interval optimization system based on real-time passenger data analysis, which solves the problems mentioned in the background section.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A rail transit interval optimization system based on real-time passenger data analysis includes: The data acquisition module acquires at least one rail vehicle and at least one station in the rail transit system, comprehensively identifies the waiting area within the station, and acquires the number of people waiting in the waiting area. The trend prediction module analyzes the existing trend of the number of people waiting in the waiting area and, based on the existing trend, obtains a waiting prediction function to predict the future number of people waiting. The data analysis module sets a preset time, takes the maximum number of people that the waiting area can accommodate as the target upper limit, and predicts the estimated number of people waiting in the waiting area after the preset time according to the waiting prediction function. The station classification module identifies stations in the waiting area where the estimated number of people waiting exceeds the target limit as feature stations, and the remaining stations as non-feature stations. An optimization simulation module is used to generate at least one time interval scheme. The module estimates the number of people in the waiting area after the time interval scheme has been implemented for a preset time, and obtains the estimated number of people after optimization. The scheme selection module selects a time interval scheme as a preliminary time interval scheme, which satisfies the requirement that the estimated number of passengers in the waiting area within the station does not exceed the target limit after optimization according to the preliminary time interval scheme. Based on the characteristic station, the target time interval scheme is selected from at least one preliminary time interval scheme, and the train interval is optimized using the target time interval scheme. The train interval optimization is completed by adjusting the train speed.

[0005] Preferably, the step of comprehensively identifying the waiting area within the station and obtaining the number of people waiting in the waiting area includes the following steps: The track on the left side of the waiting area is designated as the forward track, and the track on the right side of the waiting area is designated as the reverse track. The location of the exit within the station is obtained, and at least one continuous image of the waiting area within the station is obtained. The head contour of the passenger is identified in the continuous image. The at least one continuous image is overlaid and the position of the head contour is marked. The positions of the same head contour are connected in chronological order to obtain the movement trajectory of the head contour. If the end of the movement trajectory of the head contour is the exit within the station, then the head contour is regarded as a non-target contour; otherwise, the head contour is regarded as the target contour. If the face of the target contour is facing the positive track, then the target contour is regarded as the positive contour; otherwise, the target contour is regarded as the negative contour. The number of forward contours in the waiting area is taken as the number of forward passengers, and the number of reverse contours in the waiting area is taken as the number of reverse passengers. The forward and reverse passenger numbers are added together to obtain the number of passengers waiting in the waiting area.

[0006] Preferably, the analysis to obtain the existing trend of the number of people waiting in the waiting area includes the following steps: The time interval from the current day to the present moment is taken as the feature interval. At least one identification point is evenly selected within the feature interval. At the moment when the value of the identification point is equal to the value of the identification point, the number of people in the waiting area in the forward direction is obtained as the positive reference value of the identification point. The number of people in the reverse direction in the waiting area is obtained as the negative reference value of the identification point. The positive reference values ​​and negative reference values ​​of all identification points are summarized as the existing trend.

[0007] Preferably, the step of analyzing and obtaining the waiting number prediction function based on existing trends includes the following steps: Obtain sample data from historical data. The interval time of rail vehicles in the sample data is consistent with the current interval time of rail vehicles. Divide the time of day evenly into at least one sampling point. At the time when the value of the sampling point is equal, count the number of people in the waiting area of ​​the sample data to obtain the sample number. The sample function is obtained by pairing the values ​​of the sample points with the number of people in the sample and fitting the data. Substitute the values ​​of the identification points into the sample function to obtain the baseline values. Take the average of at least one baseline value to obtain the baseline average value. Take the average of at least one positive reference value to obtain the positive reference average value. Take the average of at least one negative reference value to obtain the negative reference average value. The positive coefficient is obtained by dividing the average of the positive reference values ​​by the base value and taking the average of the negative reference values. The positive coefficient is multiplied by the sample function to obtain the positive prediction function, and the negative coefficient is multiplied by the sample function to obtain the negative prediction function. Both the positive and negative prediction functions are used as waiting area prediction functions, and the waiting area corresponds to the positive and negative prediction functions generated by them.

[0008] Preferably, the formation of the preset time includes the following steps: In the historical data, at least one historical case of adjusting the interval between train services of rail transit is obtained. In the historical case, the time taken when the number of people waiting in the waiting area first decreased after the adjustment is obtained as the historical time. The maximum value of at least one historical time is used as the preset time.

[0009] Preferably, the step of predicting the estimated number of people waiting in the waiting area after a preset time based on the waiting prediction function includes the following steps: Substitute the time after the preset time into the forward prediction function corresponding to the waiting area to obtain the forward estimated number of people. Substitute the time after the preset time into the reverse prediction function corresponding to the waiting area to obtain the reverse estimated number of people. Add the forward estimated number of people and the reverse estimated number of people to obtain the estimated number of people waiting for the train.

[0010] Preferably, the scheme for forming at least one time interval includes the following steps: Obtain at least one normal operating condition from historical data. In the normal operating condition, obtain the lower limit of the interval time between adjacent rail vehicles as a safety value. Obtain the maximum interval time between currently adjacent rail vehicles, use it as a feature value, and use the safety value and the feature value as endpoints to form an interval range; At least one interval point is uniformly selected in the interval interval, and at least one interval point is randomly assigned to a rail vehicle. The interval point may be repeatedly assigned. Each allocation method of the interval points forms a time interval scheme in which the interval time between adjacent rail vehicles is equal to the value of the next assigned interval point among the adjacent rail vehicles.

[0011] Preferably, estimating the number of people in the waiting area after the time interval scheme has been implemented for a preset time includes the following steps: The rail vehicles that would pass through the station where the waiting area is located within the preset time when the time interval scheme is not implemented are used as target vehicles; The target vehicle running on the forward track is designated as the first forward vehicle, and the target vehicle running on the reverse track is designated as the first reverse vehicle. The rail vehicles that will pass through the station where the waiting area is located within a preset time after the implementation of the time interval scheme will be used as characteristic vehicles. A feature vehicle that will run on the forward track will be designated as a second forward vehicle, and a feature vehicle that will run on the reverse track will be designated as a second reverse vehicle. The number of the second forward-moving vehicles divided by the number of the first forward-moving vehicles gives the positive action value, and the number of the second reverse-moving vehicles divided by the number of the first reverse-moving vehicles gives the negative action value. Dividing the positively estimated number of passengers in the waiting area by the positive action value yields the positively optimized number of passengers in the waiting area. Dividing the negatively estimated number of passengers in the waiting area by the negative action value yields the negatively optimized number of passengers in the waiting area. The positively optimized number of passengers and the negatively optimized number of passengers in the waiting area are added together to obtain the positively optimized number of passengers in the waiting area.

[0012] Preferably, the step of selecting the target time interval scheme from at least one preparatory time interval scheme includes the following steps: Subtract the estimated number of passengers in the waiting area of ​​the feature station from the target upper limit to obtain the excess adjustment range of the feature station. Add up at least one excess adjustment range to obtain the total excess range. Obtain the original time interval of the rail vehicles before the implementation of the preparatory time interval plan. The original time interval is the time interval between the rail vehicle and the previous rail vehicle in the direction of travel. The difference between the original time interval of the rail vehicle and the interval point allocated to the rail vehicle in the preparatory time interval scheme is taken as the absolute value to obtain the adjustment value of the rail vehicle. At least one adjustment value is accumulated to obtain the total adjustment value. Using the analytic hierarchy process (AHP), the weights of the total adjustment value and the excess total amplitude are obtained. The total adjustment value and the excess total amplitude are multiplied by their corresponding weights and then summed to obtain the judgment value of the preparatory time interval scheme. The preparatory time interval scheme with the smallest judgment value is taken as the target time interval scheme.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: By setting up data acquisition, trend prediction, data analysis, optimization simulation, and scheme selection modules, the number of people waiting in the waiting area is collected, thereby obtaining the current trend of the number of people waiting. Combined with historical passenger flow distribution, the number of people waiting in the future can be predicted more accurately. Adjustments can then be made based on the prediction results, allowing for the selection of a more reasonable scheme from multiple options. This ensures that passenger flow demand is met while minimizing adjustments to the intervals between rail transit services. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the rail transit interval optimization system based on real-time passenger data analysis according to the present invention. Figure 2 This is a flowchart illustrating the process of comprehensively identifying the waiting area within a station and obtaining the number of people waiting in the waiting area according to the present invention. Figure 3 This is a flowchart illustrating the waiting prediction function of the present invention, which analyzes existing trends to predict the number of people waiting for trains in the future. Figure 4 This is a schematic diagram of the process for forming at least one time interval scheme according to the present invention; Figure 5 This is a flowchart illustrating the process of estimating the number of people in the waiting area after the time interval scheme of the present invention has been implemented for a preset time. Figure 6 This is a schematic diagram of the process of selecting a target time interval scheme from at least one preparatory time interval scheme according to the present invention. Detailed Implementation

[0015] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0016] Reference Figure 1 As shown, the rail transit interval optimization system based on real-time passenger data analysis includes: The data acquisition module acquires at least one rail vehicle and at least one station in the rail transit system, comprehensively identifies the waiting area within the station, and acquires the number of people waiting in the waiting area. The trend prediction module analyzes the existing trend of the number of people waiting in the waiting area and, based on the existing trend, obtains a waiting prediction function to predict the future number of people waiting. The data analysis module sets a preset time, takes the maximum number of people that the waiting area can accommodate as the target upper limit, and predicts the estimated number of people waiting in the waiting area after the preset time according to the waiting prediction function. The station classification module identifies stations in the waiting area where the estimated number of people waiting exceeds the target limit as feature stations, and the remaining stations as non-feature stations. An optimization simulation module is used to generate at least one time interval scheme. The module estimates the number of people in the waiting area after the time interval scheme has been implemented for a preset time, and obtains the estimated number of people after optimization. The scheme selection module selects a time interval scheme as a preliminary time interval scheme, which satisfies the requirement that the estimated number of passengers in the waiting area within the station does not exceed the target limit after optimization according to the preliminary time interval scheme. Based on the characteristic station, the target time interval scheme is selected from at least one preliminary time interval scheme, and the train interval is optimized using the target time interval scheme. The train interval optimization is completed by adjusting the train speed.

[0017] When optimizing train schedules, it's crucial to accurately track the number of passengers waiting in the waiting area. This allows for recording the number of passengers, revealing current trends, and predicting future trends. This solution primarily optimizes subway train intervals. Passengers enter the subway by swiping their cards, which allows for data collection, but it cannot capture those waiting in line. Regardless of the number of people entering the subway station, the trains are likely to be able to accommodate them all, preventing overcrowding. Therefore, the number of passengers obtained through card swiping is invalid and requires further analysis. The number of people waiting in the waiting area is mainly obtained through image recognition. However, there are also interfering factors, such as people getting off the rail vehicles. Therefore, corresponding steps will be set up in the subsequent steps to obtain the number of people waiting. On the other hand, each station is bidirectional, that is, there are two tracks running in opposite directions, i.e., going back and forth. When a vehicle reaches the end of one track, it will move to the other track by existing lane changing methods to move in the opposite direction. Therefore, when performing people recognition, it is also necessary to consider the two directions separately. Corresponding steps will also be set up in the subsequent steps to handle this. In addition, when optimizing the interval between train services, the main issue is the overcrowding at some stations. While significant adjustments can achieve this goal, the process is complicated and time-consuming because it is done through vehicle speed control and departure interval control after reaching the terminal station. Therefore, a solution that meets the above requirements needs to be developed by setting up corresponding steps in the subsequent steps.

[0018] Reference Figure 2 As shown, the process of comprehensively identifying the waiting area within the station and obtaining the number of people waiting in the waiting area includes the following steps: The track on the left side of the waiting area is designated as the forward track, and the track on the right side of the waiting area is designated as the reverse track. The location of the exit within the station is obtained, and at least one continuous image of the waiting area within the station is obtained. The head contour of the passenger is identified in the continuous image. The at least one continuous image is overlaid and the position of the head contour is marked. The positions of the same head contour are connected in chronological order to obtain the movement trajectory of the head contour. If the end of the movement trajectory of the head contour is the exit within the station, then the head contour is regarded as a non-target contour; otherwise, the head contour is regarded as the target contour. If the face of the target contour is facing the positive track, then the target contour is regarded as the positive contour; otherwise, the target contour is regarded as the negative contour. The number of forward contours in the waiting area is taken as the number of forward passengers, and the number of reverse contours in the waiting area is taken as the number of reverse passengers. The forward and reverse passenger numbers are added together to obtain the number of passengers waiting in the waiting area.

[0019] Image recognition is accomplished using existing technologies. Many algorithms exist for head and face recognition. During recognition, it's necessary to distinguish the number of people. In rail transit, passengers in the waiting area of ​​each station may be riding in the forward or reverse direction of a train. Therefore, when predicting the number of passengers, this distinction is crucial. This requires identifying passengers riding in the forward-moving and reverse-moving tracks. During recognition, people getting on and off the trains need to be removed, as their movement trajectories inevitably end at the station exits. Therefore, irrelevant personnel can be eliminated. Based on waiting habits, passengers on the forward-moving tracks generally face the forward track, and passengers on the reverse-moving tracks generally face the reverse track, with very few exceptions. However, the error caused by these exceptions is small and can be ignored. Therefore, the number of people traveling in the forward and reverse directions can be obtained. This method can then be used to identify the number of people waiting in the waiting area at any given time.

[0020] The analysis of existing trends in the number of people waiting in the waiting area includes the following steps: The time interval from the current day to the present moment is taken as the feature interval. At least one identification point is evenly selected within the feature interval. At the moment when the value of the identification point is equal to the value of the identification point, the number of people in the waiting area in the forward direction is obtained as the positive reference value of the identification point. The number of people in the reverse direction in the waiting area is obtained as the negative reference value of the identification point. The positive reference values ​​and negative reference values ​​of all identification points are summarized as the existing trend.

[0021] Reference Figure 3 As shown, based on existing trends, the waiting time prediction function for predicting future waiting numbers includes the following steps: Obtain sample data from historical data. The interval time of rail vehicles in the sample data is consistent with the current interval time of rail vehicles. Divide the time of day evenly into at least one sampling point. At the time when the value of the sampling point is equal, count the number of people in the waiting area of ​​the sample data to obtain the sample number. The sample function is obtained by pairing the values ​​of the sample points with the number of people in the sample and fitting the data. Substitute the values ​​of the identification points into the sample function to obtain the baseline values. Take the average of at least one baseline value to obtain the baseline average value. Take the average of at least one positive reference value to obtain the positive reference average value. Take the average of at least one negative reference value to obtain the negative reference average value. The positive coefficient is obtained by dividing the average of the positive reference values ​​by the base value and taking the average of the negative reference values. The positive coefficient is multiplied by the sample function to obtain the positive prediction function, and the negative coefficient is multiplied by the sample function to obtain the negative prediction function. Both the positive and negative prediction functions are used as waiting area prediction functions, and the waiting area corresponds to the positive and negative prediction functions generated by them.

[0022] Generally speaking, the passenger distribution ratio in a day is approximately consistent, except that the total number of people may be higher or lower on a certain day. Based on the distribution of a certain day, all waiting prediction functions can be inferred. Here, we choose the sample situation as a reference. Since the data for the entire day in the sample situation is known, the sample function can be obtained. The sample function can calculate the number of people in the waiting area at each time of day in the sample situation. Since the distribution pattern of passengers going forward or backward at each station can also be estimated using the sample function, however, since the overall passenger flow varies on a given day, the sample function needs to be adjusted proportionally. Thus, the forward prediction function and the backward prediction function corresponding to the waiting area are obtained. This proportion is obtained by comparing the average of the existing trend with the average of the sample situation. The forward and backward prediction functions here are obtained under the condition that the rail vehicles transport passengers at the current intervals. They are not the cumulative number of passengers, but the number of passengers who are stranded after transportation and those waiting to enter the station. Historical data and sample data were obtained under the condition that the rail vehicles were operating at the current intervals, because the intervals of rail transit services usually do not change, and therefore, they are consistent when not adjusted.

[0023] The process of setting a preset time includes the following steps: In the historical data, at least one historical case of adjusting the interval between train services of rail transit is obtained. In the historical case, the time taken when the number of people waiting in the waiting area first decreased after the adjustment is obtained as the historical time. The maximum value of at least one historical time is used as the preset time.

[0024] During the adjustment process, the number of waiting passengers does not change immediately upon completion of the adjustment. Therefore, it is necessary to set a preset time. The preset time is used to adjust in advance. By adjusting in advance, we can optimize the situation based on the predictions and ensure that the situation is alleviated after the preset time. Since the optimization starts at 5 or 6 in the morning, there will not be too many passengers at that time and for a period of time thereafter. The first occurrence of too many passengers will inevitably happen after the preset time. Therefore, we can make optimizations in advance. As long as we continuously optimize the situation after the preset time, we can ensure that the number of passengers waiting throughout the day is within the target limit.

[0025] Based on the waiting area prediction function, the estimated number of people waiting in the waiting area after a preset time is predicted through the following steps: Substitute the time after the preset time into the forward prediction function corresponding to the waiting area to obtain the forward estimated number of people. Substitute the time after the preset time into the reverse prediction function corresponding to the waiting area to obtain the reverse estimated number of people. Add the forward estimated number of people and the reverse estimated number of people to obtain the estimated number of people waiting for the train.

[0026] When making estimates, it is necessary to predict the number of people in the two directions separately, because the number of people waiting may differ significantly and needs to be optimized accordingly.

[0027] Reference Figure 4 As shown, forming at least one time interval scheme includes the following steps: Obtain at least one normal operating condition from historical data. In the normal operating condition, obtain the lower limit of the interval time between adjacent rail vehicles as a safety value. Obtain the maximum interval time between currently adjacent rail vehicles, use it as a feature value, and use the safety value and the feature value as endpoints to form an interval range; At least one interval point is uniformly selected in the interval interval, and at least one interval point is randomly assigned to a rail vehicle. The interval point may be repeatedly assigned. Each allocation method of the interval points forms a time interval scheme in which the interval time between adjacent rail vehicles is equal to the value of the next assigned interval point among the adjacent rail vehicles.

[0028] Reference Figure 5 As shown, estimating the number of people in the waiting area after the preset time interval scheme is implemented includes the following steps: The rail vehicles that would pass through the station where the waiting area is located within the preset time when the time interval scheme is not implemented are used as target vehicles; The target vehicle running on the forward track is designated as the first forward vehicle, and the target vehicle running on the reverse track is designated as the first reverse vehicle. The rail vehicles that will pass through the station where the waiting area is located within a preset time after the implementation of the time interval scheme will be used as characteristic vehicles. A feature vehicle that will run on the forward track will be designated as a second forward vehicle, and a feature vehicle that will run on the reverse track will be designated as a second reverse vehicle. The number of the second forward-moving vehicles divided by the number of the first forward-moving vehicles gives the positive action value, and the number of the second reverse-moving vehicles divided by the number of the first reverse-moving vehicles gives the negative action value. Dividing the positively estimated number of passengers in the waiting area by the positive action value yields the positively optimized number of passengers in the waiting area. Dividing the negatively estimated number of passengers in the waiting area by the negative action value yields the negatively optimized number of passengers in the waiting area. The positively optimized number of passengers and the negatively optimized number of passengers in the waiting area are added together to obtain the positively optimized number of passengers in the waiting area.

[0029] The more trains pass through a station, the fewer passengers will be stranded and waiting, meaning the two are inversely proportional. Since the positive estimated number of passengers is the estimated number of passengers under the operation of the first positive train, it can be assumed that the product of the positive estimated number of passengers and the number of the first positive train is equal to the product of the number of the second positive train and the number of passengers after positive optimization. Thus, the number of passengers after positive optimization can be calculated. Similarly, the number of passengers after negative optimization can also be calculated. Since the time interval between adjacent vehicles is fixed under each scheme, their arrival time at the station can be estimated. The interval can be calculated based on the vehicle that has already passed the station closest to the station. Since the time that vehicle has left the station is known, the arrival time of subsequent vehicles can be estimated given the known time interval.

[0030] Reference Figure 6 As shown, selecting the target time interval scheme from at least one preparatory time interval scheme includes the following steps: Subtract the estimated number of passengers in the waiting area of ​​the feature station from the target upper limit to obtain the excess adjustment range of the feature station. Add up at least one excess adjustment range to obtain the total excess range. Obtain the original time interval of the rail vehicles before the implementation of the preparatory time interval plan. The original time interval is the time interval between the rail vehicle and the previous rail vehicle in the direction of travel. The difference between the original time interval of the rail vehicle and the interval point allocated to the rail vehicle in the preparatory time interval scheme is taken as the absolute value to obtain the adjustment value of the rail vehicle. At least one adjustment value is accumulated to obtain the total adjustment value. Using the analytic hierarchy process (AHP), the weights of the total adjustment value and the excess total amplitude are obtained. The total adjustment value and the excess total amplitude are multiplied by their corresponding weights and then summed to obtain the judgment value of the preparatory time interval scheme. The preparatory time interval scheme with the smallest judgment value is taken as the target time interval scheme.

[0031] When making adjustments, the number of people at the featured stations should be kept just below the target limit. As long as the number does not exceed the target limit, the passengers are all within the operating load and will not cause serious passenger congestion. In addition, the adjustment range should be as small as possible. Therefore, by combining the two factors, a judgment value can be obtained, and based on this, the solution can be further screened.

[0032] Furthermore, this solution also proposes a storage medium on which a computer-readable program is stored. When the computer-readable program is invoked, it executes the aforementioned rail transit interval optimization system based on real-time passenger data analysis.

[0033] It is understandable that the storage medium can be a magnetic medium, such as a floppy disk, hard disk, or magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid-state drive (SSD).

[0034] In summary, the advantages of this invention are as follows: by setting up a data acquisition module, a trend prediction module, a data analysis module, an optimization simulation module, and a scheme selection module, the number of people waiting in the waiting area can be acquired, thereby obtaining the existing trend of the number of people waiting. Combined with historical passenger flow distribution, the number of people waiting in the future can be predicted more accurately. Adjustments can then be made based on the prediction results, thereby selecting a more reasonable scheme from multiple options. This allows for meeting passenger flow demand while minimizing adjustments to the intervals between rail transit services.

[0035] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A rail transit interval optimization system based on real-time passenger data analysis, characterized in that, include: The data acquisition module acquires at least one rail vehicle and at least one station in the rail transit system, comprehensively identifies the waiting area within the station, and acquires the number of people waiting in the waiting area. The trend prediction module analyzes the existing trend of the number of people waiting in the waiting area and, based on the existing trend, obtains a waiting prediction function to predict the future number of people waiting. The data analysis module sets a preset time, takes the maximum number of people that the waiting area can accommodate as the target upper limit, and predicts the estimated number of people waiting in the waiting area after the preset time according to the waiting prediction function. The station classification module identifies stations in the waiting area where the estimated number of people waiting exceeds the target limit as feature stations, and the remaining stations as non-feature stations. An optimization simulation module is used to generate at least one time interval scheme. The number of people in the waiting area after the time interval scheme has been implemented for a preset time is estimated to obtain the estimated number of people after optimization. The scheme selection module selects a time interval scheme as a preliminary time interval scheme, which satisfies the requirement that the estimated number of passengers in the waiting area within the station does not exceed the target limit after optimization according to the preliminary time interval scheme. Based on the characteristic station, the target time interval scheme is selected from at least one preliminary time interval scheme, and the train interval is optimized using the target time interval scheme. The train interval optimization is completed by adjusting the train speed.

2. The rail transit interval optimization system based on real-time passenger data analysis according to claim 1, characterized in that, The process of comprehensively identifying the waiting area within the station and obtaining the number of people waiting in the waiting area includes the following steps: The track on the left side of the waiting area is designated as the forward track, and the track on the right side of the waiting area is designated as the reverse track. The location of the exit within the station is obtained, and at least one continuous image of the waiting area within the station is obtained. The head contour of the passenger is identified in the continuous image. The at least one continuous image is overlaid and the position of the head contour is marked. The positions of the same head contour are connected in chronological order to obtain the movement trajectory of the head contour. If the end of the movement trajectory of the head contour is the exit within the station, then the head contour is regarded as a non-target contour; otherwise, the head contour is regarded as the target contour. If the face of the target contour is facing the positive track, then the target contour is regarded as the positive contour; otherwise, the target contour is regarded as the negative contour. The number of forward contours in the waiting area is taken as the number of forward passengers, and the number of reverse contours in the waiting area is taken as the number of reverse passengers. The forward and reverse passenger numbers are added together to obtain the number of passengers waiting in the waiting area.

3. The rail transit interval optimization system based on real-time passenger data analysis according to claim 2, characterized in that, The analysis revealed the existing trends in the number of people waiting in the waiting area, including the following steps: The time interval from the current day to the present moment is taken as the feature interval. At least one identification point is evenly selected within the feature interval. At the moment when the value of the identification point is equal to the value of the identification point, the number of people in the waiting area in the forward direction is obtained as the positive reference value of the identification point. The number of people in the reverse direction in the waiting area is obtained as the negative reference value of the identification point. The positive reference values ​​and negative reference values ​​of all identification points are summarized as the existing trend.

4. The rail transit interval optimization system based on real-time passenger data analysis according to claim 3, characterized in that, The process of analyzing existing trends to obtain a prediction function for future waiting numbers includes the following steps: Obtain sample data from historical data. The interval time of rail vehicles in the sample data is consistent with the current interval time of rail vehicles. Divide the time of day evenly into at least one sampling point. At the time when the value of the sampling point is equal, count the number of people in the waiting area of ​​the sample data to obtain the sample number. The sample function is obtained by pairing the values ​​of the sample points with the number of people in the sample and fitting the data. Substitute the values ​​of the identification points into the sample function to obtain the baseline values. Take the average of at least one baseline value to obtain the baseline average value. Take the average of at least one positive reference value to obtain the positive reference average value. Take the average of at least one negative reference value to obtain the negative reference average value. The positive coefficient is obtained by dividing the average of the positive reference values ​​by the base value and taking the average of the negative reference values. The positive coefficient is multiplied by the sample function to obtain the positive prediction function, and the negative coefficient is multiplied by the sample function to obtain the negative prediction function. Both the positive and negative prediction functions are used as waiting area prediction functions, and the waiting area corresponds to the positive and negative prediction functions generated by them.

5. The rail transit interval optimization system based on real-time passenger data analysis according to claim 4, characterized in that, The process of setting the preset time includes the following steps: In the historical data, at least one historical case of adjusting the interval between train services of rail transit is obtained. In the historical case, the time taken when the number of people waiting in the waiting area first decreased after the adjustment is obtained as the historical time. The maximum value of at least one historical time is used as the preset time.

6. The rail transit interval optimization system based on real-time passenger data analysis according to claim 5, characterized in that, The step of predicting the estimated number of people waiting in the waiting area after a preset time based on the waiting prediction function includes the following steps: Substitute the time after the preset time into the forward prediction function corresponding to the waiting area to obtain the forward estimated number of people. Substitute the time after the preset time into the reverse prediction function corresponding to the waiting area to obtain the reverse estimated number of people. Add the forward estimated number of people and the reverse estimated number of people to obtain the estimated number of people waiting for the train.

7. The rail transit interval optimization system based on real-time passenger data analysis according to claim 6, characterized in that, The scheme for forming at least one time interval includes the following steps: Obtain at least one normal operating condition from historical data. In the normal operating condition, obtain the lower limit of the interval time between adjacent rail vehicles as a safety value. Obtain the maximum interval time between currently adjacent rail vehicles, use it as a feature value, and use the safety value and the feature value as endpoints to form an interval range; At least one interval point is uniformly selected in the interval interval, and at least one interval point is randomly assigned to a rail vehicle. The interval point may be repeatedly assigned. Each allocation method of the interval points forms a time interval scheme in which the interval time between adjacent rail vehicles is equal to the value of the next assigned interval point among the adjacent rail vehicles.

8. The rail transit interval optimization system based on real-time passenger data analysis according to claim 7, characterized in that, The estimation of the number of people in the waiting area after the time interval scheme has been implemented for a preset time includes the following steps: The rail vehicles that would pass through the station where the waiting area is located within the preset time when the time interval scheme is not implemented are used as target vehicles; The target vehicle running on the forward track is designated as the first forward vehicle, and the target vehicle running on the reverse track is designated as the first reverse vehicle. The rail vehicles that will pass through the station where the waiting area is located within a preset time after the implementation of the time interval scheme will be used as characteristic vehicles. A feature vehicle that will run on the forward track will be designated as a second forward vehicle, and a feature vehicle that will run on the reverse track will be designated as a second reverse vehicle. The number of the second forward-moving vehicles divided by the number of the first forward-moving vehicles gives the positive action value, and the number of the second reverse-moving vehicles divided by the number of the first reverse-moving vehicles gives the negative action value. Dividing the positively estimated number of passengers in the waiting area by the positive action value yields the positively optimized number of passengers in the waiting area. Dividing the negatively estimated number of passengers in the waiting area by the negative action value yields the negatively optimized number of passengers in the waiting area. The positively optimized number of passengers and the negatively optimized number of passengers in the waiting area are added together to obtain the positively optimized number of passengers in the waiting area.

9. The rail transit interval optimization system based on real-time passenger data analysis according to claim 8, characterized in that, The process of selecting the target time interval scheme from at least one preparatory time interval scheme includes the following steps: Subtract the estimated number of passengers in the waiting area of ​​the feature station from the target upper limit to obtain the excess adjustment range of the feature station. Add up at least one excess adjustment range to obtain the total excess range. Obtain the original time interval of the rail vehicles before the implementation of the preparatory time interval plan. The original time interval is the time interval between the rail vehicle and the previous rail vehicle in the direction of travel. The difference between the original time interval of the rail vehicle and the interval point allocated to the rail vehicle in the preparatory time interval scheme is taken as the absolute value to obtain the adjustment value of the rail vehicle. At least one adjustment value is accumulated to obtain the total adjustment value. Using the analytic hierarchy process (AHP), the weights of the total adjustment value and the excess total amplitude are obtained. The total adjustment value and the excess total amplitude are multiplied by their corresponding weights and then summed to obtain the judgment value of the preparatory time interval scheme. The preparatory time interval scheme with the smallest judgment value is taken as the target time interval scheme.

Images