Public transportation capacity monitoring and dispatching method and system

Abstract

The present invention belongs to the technical field of public transportation scheduling, and discloses a method and system for monitoring and scheduling public transportation capacity. The method collects real-time passenger flow, historical passenger flow and environmental factor data of target bus routes, and performs standardization and anomaly detection; constructs a hybrid prediction model including a time series decomposition model and a random fluctuation model to predict passenger flow in future time periods; divides vehicles into fixed shifts, mobile shifts and standby shifts, and dynamically adjusts vehicle configuration according to predicted full load factors; and realizes accurate monitoring and dynamic scheduling of public transportation capacity through multi-source data fusion and intelligent prediction algorithms, thereby improving vehicle utilization and service quality and reducing operating costs.

Description

Technical Field

[0001] The present invention belongs to the technical field of public transportation scheduling, and in particular relates to a method and system for monitoring and scheduling public transportation capacity. Background Art

[0002] Traditional public transportation scheduling methods mainly rely on static departure schedules and empirical capacity arrangements, and lack the ability to respond to changes in passenger flow in real time.

[0003] Most existing public transportation dispatch systems lack the ability to accurately predict passenger flow. Traditional dispatch systems rely primarily on historical data or simple statistical models, which fail to effectively capture the impact of external factors such as weather changes, holidays, and major events on passenger flow. This results in slow response to sudden changes in passenger flow and untimely dispatch. This is particularly true during peak hours, unusual weather conditions, or major events, where insufficient or excessive vehicle allocation often occurs, resulting in wasted resources and reduced service quality. Vehicles mostly operate on fixed schedules, making it difficult to dynamically adjust to actual passenger flow. This one-size-fits-all dispatch approach is inefficient in urban environments with large passenger flow fluctuations, failing to meet passenger demand during peak periods while wasting resources during off-peak periods. These shortcomings lead to inefficient public transportation resource allocation, a poor passenger experience, increased operating costs, and hinder the sustainable development of urban public transportation systems. Patent CN114091757B discloses an intelligent scheduling method for buses and subways based on passenger flow analysis and prediction. The method evenly divides the day T into multiple time periods, and counts the historical public transportation passenger flow data for each time period of the day T; inputs the historical public transportation passenger flow data for each time period of the day T into a prediction model to obtain the public transportation predicted passenger flow data for each time period of day T+1; obtains the total public transportation predicted passenger flow N1 of station N in the time period (t1, t2), and the time period (t1, t2) includes one or more time periods; obtains the capacity N2 of station N in the time period (t1, t2); S5, performs capacity matching; increases or decreases the capacity of station N according to the p value; this method relies solely on historical passenger flow data for prediction, ignores the significant impact of environmental factors on passenger flow, resulting in limited prediction accuracy, especially during weather changes, holidays or special events when the prediction deviation is large.

[0004] Therefore, there is an urgent need for an intelligent public transportation capacity monitoring and scheduling method and system that can achieve accurate passenger flow prediction, multi-source data fusion and dynamic capacity scheduling. Summary of the Invention

[0005] The present invention provides a public transportation capacity monitoring and scheduling method and system, which are used to solve the problem that traditional public transportation scheduling lacks real-time response capability and accurate passenger flow prediction, and realize the purpose of intelligent and dynamic configuration and scheduling of public transportation capacity resources.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a method for monitoring and dispatching public transportation capacity, the method comprising:

[0008] Step S1, standardizing and detecting anomalies in the collected data of the target bus route to generate preprocessed data, wherein the collected data includes real-time passenger flow, historical passenger flow, and environmental factor data of the target bus route;

[0009] After standardization and anomaly detection of the collected data, the data is divided into an operational dataset RT, a historical dataset HS, and an environmental factor dataset EF. The operational dataset RT records real-time passenger flow, vehicle operation frequency, and full load factor; the historical dataset HS contains passenger flow statistics for the last 30 days; and the environmental factor dataset EF contains weather data and holiday information.

[0010] The standardization adopts the maximum and minimum normalization method. , is the original data, and are the maximum and minimum values ​​of the data, respectively. is the standardized data;

[0011] The anomaly detection adopts the interval range method to eliminate the abnormality that exceeds Range data, where is the data mean, is the data standard deviation;

[0012] Step S2: Based on the preprocessed data, the passenger flow of the target bus line in the future period is predicted using the constructed hybrid prediction model MM to obtain the predicted passenger flow. The hybrid prediction model MM includes a time series decomposition model TM and a random fluctuation model RM; the time series decomposition model TM decomposes the passenger flow into basic passenger flow, periodic passenger flow and trend passenger flow; the random fluctuation model RM captures short-term random fluctuations.

[0013] Step S3: The vehicles to be dispatched are divided into fixed shifts FT, flexible shifts AT and standby shifts BT, and the predicted full load factor of the target bus line is obtained according to the predicted passenger flow.

[0014] Based on the predicted full load factor, the vehicle configuration is dynamically adjusted using transport scheduling rules.

[0015] Furthermore, the expression of the time series decomposition model TM is: ;in, is the total passenger flow at time t predicted; basic passenger flow Indicates a fixed minimum passenger flow level, which is the average of the lowest passenger flow in the same period of history; Periodic passenger flow Indicates the cyclical passenger flow changes obtained by daily and weekly statistics:

[0016] ,in and are the daily cycle amplitude and phase, and are the periodic amplitude and phase, respectively;

[0017] Trending passenger flow Indicates the long-term trend, calculated by the moving average method; Indicates short-term random fluctuations.

[0018] Furthermore, the random fluctuation model RM is a model established based on environmental factor data, and the environmental factor data includes: weather influencing factors and holiday impact factors .

[0019] Weather influencing factors The value is 1.0 for non-rainy or snowy weather, 0.8 for heavy rain or rainstorm, 0.7 for snowy days, and 1.2 for other periods; Holiday impact factor , the value is 1.0 for weekdays, 1.1-1.3 for ordinary weekends, and 1.3-1.5 for long holidays; the calculation formula of the stochastic fluctuation model RM is: ,in, is the standard deviation of passenger flow in the same historical period, is a random variable that follows a standard normal distribution.

[0020] Furthermore, the load factor is predicted The calculation formula is: ,in, is the number of running vehicles at time t, The rated passenger capacity of a bicycle.

[0021] Furthermore, the vehicles are divided into fixed shifts (FT), flexible shifts (AT), and backup shifts (BT). 60% of the fixed shifts (FT) and 40% of the flexible shifts (AT) are used for vehicle scheduling. The capacity scheduling is based on the following decision rules:

[0022] When the load factor is predicted When the number of mobile shifts AT is reduced, the proportion of reduced mobile shifts AT is , but the maximum reduction rate of mobile shift AT shall not exceed 40%;

[0023] When the load factor is predicted When maintaining the current capacity configuration;

[0024] When the load factor is predicted When the number of mobile shifts AT is increased, the proportion of increased mobile shifts AT is ;

[0025] When the load factor is predicted When the emergency occurs, all standby BT shifts will be activated and vehicles from nearby lines will be dispatched for support.

[0026] Furthermore, the hybrid prediction model MM is evaluated by the mean absolute percentage error , root mean square error, and coefficient of determination are evaluated:

[0027] Mean absolute percentage error: ;

[0028] Root mean square error ;

[0029] Coefficient of determination ;

[0030] when or When the model parameter adjustment mechanism is triggered, the basic passenger flow is recalculated. and periodic passenger flow , update the hybrid prediction model MM.

[0031] Furthermore, the predicted passenger flow is obtained by using a multi-time-scale sliding window for passenger flow prediction, and the multi-time-scale sliding window includes: a short-term prediction window, a medium-term prediction window and a long-term prediction window. The short-term prediction window SW is used to predict passenger flow changes in the next 30-60 minutes, with a sampling frequency of 5 minutes; the medium-term prediction window MW is used to predict passenger flow changes in the next 3-6 hours, with a sampling frequency of 15 minutes; the long-term prediction window LW is used to predict passenger flow changes in the next 24 hours, with a sampling frequency of 30 minutes.

[0032] In a second aspect, the present invention provides a public transportation capacity monitoring and dispatching system for implementing the public transportation capacity monitoring and dispatching method of the first aspect. The system includes: a data acquisition module, a passenger flow prediction module and a capacity dispatching module.

[0033] The data acquisition module is used to collect real-time passenger flow, historical passenger flow and environmental factor data of the target bus route; it is used to standardize and detect anomalies in the collected data; the passenger flow prediction module is used to establish a hybrid prediction model MM to predict passenger flow in different time periods in the future; the capacity scheduling module is used to dynamically adjust vehicle configuration according to the predicted passenger flow.

[0034] Furthermore, the data acquisition module includes: an on-board passenger flow statistics unit, which collects passenger flow data on getting on and off the vehicle through infrared sensors or weight sensors; a station passenger flow monitoring unit, which monitors the number of passengers waiting at the station through cameras and intelligent video analysis systems; a vehicle GPS positioning unit, which collects real-time position, speed and operating status information of the vehicle; and an environmental information acquisition unit, which obtains weather information. All of the above units are connected to the data processing system of the data acquisition module in real time through a wireless communication network.

[0035] Furthermore, the system also includes a visual display module and an emergency processing module;

[0036] The visual display module is used to display the line operation status, vehicle distribution, passenger flow density heat map and full load rate warning information in real time;

[0037] When the predicted full load rate exceeds 95%, the emergency handling module automatically triggers the emergency plan, including: temporarily adding shuttle buses, adjusting departure intervals, and linking with surrounding bus routes. At the same time, it pushes congestion reminders and alternative travel suggestions to passengers through the public information platform.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] By establishing a hybrid prediction model and a multi-level scheduling strategy, the present invention achieves accurate prediction of public transportation passenger flow and intelligent scheduling of transport capacity, which can significantly increase vehicle utilization by 15-20%, reduce passenger waiting time by more than 30%, and reduce operating costs by about 10-15%. It improves the reliability and resilience of the public transportation system and provides an intelligent solution for urban traffic management. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a flow chart of a method for monitoring and dispatching public transportation capacity according to the present invention;

[0041] Figure 2 This is a schematic diagram of the composition of a public transportation capacity monitoring and dispatching system of the present invention. DETAILED DESCRIPTION

[0042] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention are described clearly and completely below. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of the present invention.

[0043] Example 1: Figure 1The figure below is a flow chart of a public transportation capacity monitoring and scheduling method according to the present invention. From a program perspective, the execution of the process can be a program installed on an application server or application terminal. It is understood that the method can be executed by any device, equipment, platform, or device cluster with computing and processing capabilities.

[0044] The method comprises:

[0045] Step S1 : standardizing and detecting anomalies on the collected data of the target bus line to generate pre-processed data, wherein the collected data includes real-time passenger flow, historical passenger flow and environmental factor data of the target bus line.

[0046] After the collected data is standardized and anomaly detected, it is composed of an operating data set RT, a historical data set HS, and an environmental factor data set EF; the operating data set RT records real-time passenger flow, vehicle operation frequency, and full load rate; the historical data set HS contains passenger flow statistics for the past 30 days; and the environmental factor data set EF contains weather data and holiday information.

[0047] In this embodiment, data collection adopts a multi-source heterogeneous data fusion strategy to ensure the comprehensiveness and accuracy of the data; real-time passenger flow data is collected through both on-board sensors and station cameras and summarized every 5 minutes; historical passenger flow data includes a 30-day rolling data window, organized by hourly, daily, and weekly time dimensions; environmental factor data includes real-time weather conditions, weather forecasts for the next 48 hours, and pre-coded holiday information.

[0048] For example, the system records the number of passengers boarding and alighting at each stop every five minutes during the weekday morning rush hour (6:30 AM to 8:30 AM). The system then compares this data with the same period over the previous four weeks. Taking into account the weather conditions (such as light rain) of the day, the system calculates outliers in real time and applies replacement strategies. On the eve of a major shopping festival, the system successfully identified passenger flow increases that exceeded normal fluctuations by comparing historical data for the same period, preventing this reasonable increase from being misclassified as an anomaly.

[0049] The standardization adopts the maximum and minimum normalization method. , is the original data, and are the maximum and minimum values ​​of the data, respectively. The data are standardized.

[0050] Normalization ensures that data of different dimensions can be effectively compared and integrated. In real applications, passenger flow often fluctuates between 0 and 2,000 people / hour. By normalizing the maximum and minimum values, we convert them into standardized values ​​between 0 and 1. For example, if a station's historical maximum passenger flow is 1,800 people / hour and its minimum is 50 people / hour, and the current detected passenger flow is 1,000 people / hour, the normalized value is (1,000 - 50) / (1,800 - 50) ≈ 0.54.

[0051] The anomaly detection adopts the interval range method to eliminate the abnormality that exceeds Range data, where is the data mean, is the data standard deviation. This range covers approximately 95% of normal data, effectively identifying outliers without excessively eliminating fluctuating data. During implementation, outliers are not discarded directly but replaced with the average of the adjacent time period to maintain data continuity. For continuous anomalies, the system triggers an alert and requires manual review to prevent misidentification of normal large-scale passenger flow fluctuations (such as large-scale events) as abnormal data.

[0052] The anomaly detection component uses a sliding window mechanism to implement dynamic threshold setting. The system automatically updates the mean μ and standard deviation σ for each time period weekly to adapt to seasonal variations. For example, during the summer vacation, the average weekday passenger flow on a certain route drops to 800 passengers / hour, with a standard deviation of 100 passengers / hour. The anomaly threshold range is [600, 1000]. If a passenger flow of 1100 passengers / hour is detected, the system will not directly exclude it. Instead, it will record it as an anomaly and temporarily replace it with the average of 850 passengers / hour for the same period over the previous three days, triggering a manual review notification.

[0053] Step S2: Based on the preprocessed data, the passenger flow of the target bus line in the future period is predicted using the constructed hybrid prediction model MM to obtain the predicted passenger flow. The hybrid prediction model MM includes a time series decomposition model TM and a random fluctuation model RM; the time series decomposition model TM decomposes the passenger flow into basic passenger flow, periodic passenger flow and trend passenger flow; the random fluctuation model RM captures short-term random fluctuations.

[0054] The core advantage of the hybrid forecasting model (MM) lies in its ability to separate deterministic and stochastic components. The time series decomposition model (TM) captures regular variations, while the stochastic fluctuation model (RM) addresses short-term fluctuations. In practice, the system uses Fourier analysis to determine the daily cycle amplitude Ad and phase φd, as well as the weekly cycle amplitude Aw and phase φw. For example, analysis of three months of data for a major urban bus route revealed the following: daily cycle amplitude Ad = 400 (passengers / hour), phase φd = π / 2 (corresponding to the morning rush hour), weekly cycle amplitude Aw = 200 (passengers / hour), and phase φw = 0 (passenger flow is highest on Mondays). This means that weekday morning rush hour passenger flow is approximately 400 passengers / hour higher than the daily average, while Monday passenger flow is generally approximately 200 passengers / hour higher than on weekends. The system automatically updates these parameters monthly to ensure the model consistently reflects the latest passenger flow patterns.

[0055] The expression of the time series decomposition model TM is: ;in, is the total passenger flow at time t predicted; basic passenger flow Indicates a fixed minimum passenger flow level, which is the average of the lowest passenger flow in the same period of history; Periodic passenger flow Indicates the cyclical passenger flow changes obtained by daily and weekly statistics:

[0056] ,in and are the daily cycle amplitude and phase, and are the cycle amplitude and phase, respectively.

[0057] Trending passenger flow Indicates the long-term trend, calculated by the moving average method; Indicates short-term random fluctuations.

[0058] The random fluctuation model RM is a model established based on environmental factor data, and the environmental factor data includes: weather influencing factors and holiday impact factors , weather influencing factors The value is 1.0 for non-rainy or snowy weather, 0.8 for heavy rain or rainstorm, 0.7 for snowy days, and 1.2 for other periods; Holiday impact factor , the value for weekdays is 1.0, the value for ordinary weekends is 1.1-1.3, and the value for long holidays is 1.3-1.5; the formula for calculating random fluctuations is: ,in, is the standard deviation of passenger flow in the same historical period, is a random variable that follows a standard normal distribution.

[0059] The weather impact factor, Wf, is determined based on historical passenger flow comparisons for the same period. During heavy rain or torrential rain, passenger flow decreases by an average of 20% (hence the value of 0.8). During snowy days, passenger flow decreases even more, averaging 30% (hence the value of 0.7). On days of light rain or mild haze, public transportation ridership actually increases by approximately 20% (hence the value of 1.2). This is primarily due to restrictions on private car travel, prompting passengers to turn to public transportation. The holiday impact factor, Ef, is further refined based on the type of holiday. For typical weekends, the value ranges from 1.1 to 1.3, with commercial routes taking the upper limit of 1.3 and residential routes taking the lower limit of 1.1. During long holidays, scenic routes take a value of 1.5, commercial routes 1.4, and residential routes 1.3. For example, during the May Day holiday, bus traffic to Happy Valley is expected to increase by 50% compared to weekdays.

[0060] Other periods primarily refer to exceptional weather conditions such as light rain and haze. During these periods, passenger flow actually increases by approximately 20%, primarily due to a shift from private car travel to public transportation. The range of holiday impact factors allows for fine-tuning based on specific holiday types and local characteristics. For example, the impact of shopping festivals and traditional holidays varies. The random variable ε introduces randomness based on statistical properties, enabling the model to simulate real-world uncertainty.

[0061] Step S3: The vehicles to be dispatched are divided into fixed shifts FT, flexible shifts AT and standby shifts BT, and the predicted full load factor of the target bus line is obtained according to the predicted passenger flow.

[0062] Based on the predicted full load factor, the vehicle configuration is dynamically adjusted using transport scheduling rules.

[0063] Forecast load factor The calculation formula is: ,in, is the number of running vehicles at time t, The rated passenger capacity of a bicycle.

[0064] Vehicles are divided into fixed shifts FT, mobile shifts AT and standby shifts BT. In the public transportation capacity scheduling system, vehicles are divided into three types of shifts.

[0065] Fixed-flight FT vehicles, representing 60% of the total fleet, operate according to a fixed schedule and frequency, ensuring a stable baseline service level and resisting adjustments due to short-term passenger flow fluctuations. Flexible-flight AT vehicles, comprising approximately 40%, are flexible and can be adjusted based on passenger flow forecasts. Their increase or decrease ratio is based on a specific calculation formula. Backup-flight BT vehicles are additional vehicles deployed in special or emergency situations, serving as the bus system's last resort to handle peak passenger flow or unexpected situations. This three-tiered vehicle classification scheme maintains service stability while flexibly responding to passenger flow fluctuations, improving vehicle utilization and reducing operating costs.

[0066] Vehicle scheduling is based on 60% fixed shifts (FT) and 40% mobile shifts (AT). This ratio of 60% fixed shifts to 40% mobile shifts is the optimal balance point obtained through simulation tests on multiple routes and various passenger flow patterns. It ensures service stability while providing sufficient scheduling flexibility.

[0067] The setting of the full load rate threshold takes into account the balance between passenger comfort and operational efficiency. The lower limit of 60% is determined based on the carriage comfort research to ensure that passengers have sufficient personal space; the upper limit of 80% takes into account the rational use of transport capacity during peak periods; and the emergency threshold of 90% is to avoid a sudden drop in service quality and increased safety risks caused by extreme congestion.

[0068] Transport scheduling is based on the following decision rules:

[0069] When the load factor is predicted When the number of mobile shifts AT is reduced, the reduction ratio is , but the maximum reduction rate shall not exceed 40%;

[0070] When the load factor is predicted When maintaining the current capacity configuration;

[0071] When the load factor is predicted When the mobile shift AT is increased, the increase ratio is ;

[0072] When the load factor is predicted At this time, all backup BT shifts are activated and vehicles from nearby lines are dispatched for support.

[0073] The dispatch ratio calculation formula is designed to follow a gradual approach. For example, when the predicted full load factor is 85%, the ratio of the increase in flexible shifts is (85% - 80%) / 0.1 = 50%, which means the current flexible shifts are increased by half. This gradual adjustment avoids overreaction while ensuring that resource inputs match demand growth.

[0074] The mean absolute percentage error was used to evaluate the prediction model , root mean square error, and coefficient of determination:

[0075] Mean absolute percentage error: ;

[0076] Root mean square error ;

[0077] Coefficient of determination ;

[0078] when or When the model parameter adjustment mechanism is triggered, the basic passenger flow is recalculated. and periodic passenger flow .

[0079] The predicted passenger flow is obtained by using a multi-time scale sliding window for passenger flow prediction, and the multi-time scale sliding window includes: a short-term prediction window, a medium-term prediction window and a long-term prediction window. The short-term prediction window SW is used to predict passenger flow changes in the next 30-60 minutes, with a sampling frequency of 5 minutes; the medium-term prediction window MW is used to predict passenger flow changes in the next 3-6 hours, with a sampling frequency of 15 minutes; the long-term prediction window LW is used to predict passenger flow changes in the next 24 hours, with a sampling frequency of 30 minutes.

[0080] The short-term prediction window SW is mainly used for real-time scheduling decisions, and a 5-minute sampling frequency can promptly capture passenger flow fluctuations. The medium-term prediction window MW supports scheduling adjustments and resource allocation, and a 15-minute sampling frequency balances prediction accuracy and computational efficiency. The long-term prediction window LW is used for strategic planning and resource preparation, and a 30-minute sampling frequency is sufficient to reflect passenger flow trends within a day.

[0081] The multi-timescale strategy enables the system to simultaneously address decision-making needs across different timeframes. The short-term prediction window (SW) (5-minute sampling frequency) is primarily used for real-time scheduling decisions, such as temporarily adding vehicles. The medium-term prediction window (MW) (15-minute sampling frequency) guides on-duty personnel and vehicle scheduling adjustments. The long-term prediction window (LW) (30-minute sampling frequency) is used for next-day vehicle deployment and personnel scheduling.

[0082] The characteristic variables and computing resources used in each forecast window are also different: the short-term window SW mainly relies on real-time data of the last 30 minutes and a simplified forecast model, and the calculation delay is controlled within 10 seconds; the medium-term window MW combines the accumulated data of the day and the historical data of the same period, and the model complexity is moderate; the long-term window LW makes full use of all available information such as historical data, weather forecasts and event calendars, and adopts a complete hybrid forecast model.

[0083] When predicting passenger flow using multi-timescale sliding windows, each window uses different feature variables and computing resources. The long-term forecast window uses a complete hybrid forecasting model, while the short-term and medium-term windows use simplified versions of the hybrid forecasting model to balance prediction accuracy and computational efficiency.

[0084] The short-term prediction window SW mainly relies on the real-time data of the last 30 minutes and a simplified prediction model, and the calculation delay is controlled within 10 seconds. The hybrid prediction model expression of the short-term prediction window is:

[0085] ;in, The original definition of basic passenger flow is retained, indicating a fixed minimum passenger flow level, which is taken as the average of the lowest passenger flow in the same period in history; Considering only the periodic pattern of the last 30 minutes, it simplifies to: ;

[0086] For real-time random fluctuations, the simplified calculation is: ;in is the standard deviation of passenger flow in the last 30 minutes, is the weather influencing factor, is a random variable that follows a standard normal distribution.

[0087] The medium-term forecast window MW combines the accumulated data of the day and the historical data of the same period, and the model complexity is moderate. The hybrid forecast model expression of the medium-term forecast window is:

[0088] ;

[0089] in, The original basic passenger flow definition is retained; Consider the cyclical pattern for the day:

[0090] ; To simplify the trend passenger flow, only the moving average of the last 7 days is considered; For medium-term stochastic fluctuations: ;

[0091] in is the standard deviation of passenger flow during the same period of the day, is the weather influencing factor, is the holiday impact factor, is a random variable that follows a standard normal distribution.

[0092] This multi-time-scale hybrid prediction model design enables the system to simultaneously respond to decision-making needs in different time ranges. The short-term prediction window is mainly used for real-time scheduling decisions, the medium-term prediction window guides the scheduling of on-duty personnel and vehicles, and the long-term prediction window is used for next-day vehicle deployment and personnel arrangement planning, thereby realizing comprehensive intelligent management of public transportation capacity.

[0093] This technology has been piloted for six months on a trunk line in a certain city, achieving significant results. The line is 32 kilometers long, has 35 stations, and handles a peak one-way passenger flow of 2,200 passengers per hour. After applying this method, the system performance is as follows:

[0094] Passenger flow forecast accuracy has significantly improved, with the MAPE dropping from 23.7% to 8.5%. This improvement is particularly evident during severe weather and special events. For example, during a sudden downpour in February 2024, the system accurately predicted a 16.8% drop in passenger flow, enabling timely adjustments to capacity allocation and maintaining vehicle utilization rates above 85%.

[0095] Transport resource utilization efficiency has been significantly improved, with the average load factor rising from 62% to 76%, and complaints about congestion during peak hours decreasing by 42%. By dynamically adjusting schedules, the system deployed five additional vehicles during the peak weekday morning rush hour of 7:00-8:00 AM, effectively alleviating congestion. During the off-peak period of 10:00 AM to 3:00 PM, eight vehicles were removed from the previously scheduled fixed schedule, reducing fuel consumption and vehicle wear.

[0096] The comprehensive economic benefits are significant. The average monthly operating cost of the pilot line has dropped by 14.3%, saving about 2,800 liters of fuel per month and reducing carbon emissions by about 6.5 tons. At the same time, the average waiting time for passengers has been shortened from 12 minutes to 8.5 minutes, and the service satisfaction score has increased from 75 points to 87 points. Especially during the 2023 National Day Golden Week, the system successfully coped with the challenge of a 35% increase in passenger flow compared to weekdays through precise prediction and advance scheduling, maintained good service quality, and won unanimous praise from passengers and operators. This practical application case fully demonstrates the remarkable effect of the present invention in improving the efficiency of public transportation resource utilization, reducing operating costs and improving service quality, and provides a feasible technical solution for the intelligent management of urban public transportation.

[0097] Example 2: Figure 2 As shown, the present invention provides a schematic diagram of the composition of a public transportation capacity monitoring and scheduling system, which is used to implement the public transportation capacity monitoring and scheduling method of the first aspect. The system includes: a data acquisition module, a passenger flow prediction module and a capacity scheduling module.

[0098] The system adopts a modular design, with each functional module connected via standardized interfaces for easy maintenance and upgrades. The data acquisition module, serving as the system's foundation, provides multi-source, heterogeneous data; the passenger flow forecasting module enables scientific predictions of future passenger flows; and the capacity dispatching module, the decision-making and execution layer, translates forecasts into specific vehicle dispatch instructions. These three main modules form a closed loop of data acquisition, analysis and forecasting, and decision-making and execution, enabling intelligent management of public transportation capacity.

[0099] The data acquisition module is used to collect real-time passenger flow, historical passenger flow and environmental factor data of the target bus route; it is used to standardize and detect anomalies in the collected data; the passenger flow prediction module is used to establish a hybrid prediction model MM to predict passenger flow in different time periods in the future; the capacity scheduling module is used to dynamically adjust vehicle configuration according to the predicted passenger flow.

[0100] The data acquisition module includes: an on-board passenger flow statistics unit, which collects passenger flow data on getting on and off the vehicle through infrared sensors or weight sensors; a station passenger flow monitoring unit, which monitors the number of passengers waiting at the station through cameras and an intelligent video analysis system; a vehicle GPS positioning unit, which collects real-time vehicle location, speed and operating status information; and an environmental information collection unit, which obtains weather information. All of these units are connected to the data processing system of the data acquisition module in real time through a wireless communication network.

[0101] The data collection system utilizes a distributed architecture, with each collection unit operating independently and transmitting data in real time via a wireless network. The onboard passenger flow counting unit utilizes dual verification using infrared sensors and weight sensors, achieving an accuracy rate exceeding 95%. The station passenger flow monitoring unit utilizes video analysis technology to automatically identify the number of waiting passengers and pre-processes data through edge computing to reduce transmission overhead. The vehicle GPS positioning unit collects location data every 10 seconds, enabling precise tracking of vehicle movements. The environmental information collection unit uses an API interface to obtain real-time weather forecasts and measured data provided by the meteorological department. All collection units utilize breakpoint-resume transmission and data caching mechanisms to ensure data integrity during network fluctuations.

[0102] The system also includes a visual display module and an emergency processing module;

[0103] The visual display module is used to display the line operation status, vehicle distribution, passenger flow density heat map and full load rate warning information in real time;

[0104] When the predicted full load rate exceeds 95%, the emergency handling module automatically triggers the emergency plan, including: temporarily adding shuttle buses, adjusting departure intervals, and linking with surrounding bus routes. At the same time, it pushes congestion reminders and alternative travel suggestions to passengers through the public information platform.

[0105] The visualization module features an interactive dashboard design, supporting multi-dimensional data display and drill-down analysis. Route operation status displays include real-time vehicle location, speed, and delays. A passenger flow density heat map visually displays passenger flow distribution at each station and route section. Load factor warning information is displayed in a graded format: sections below 60% are displayed green, 60%-80% are yellow, 80%-90% are orange, and above 90% are red, allowing dispatchers to quickly identify areas requiring attention. The emergency response module incorporates a multi-level response mechanism. The 95% load factor trigger point is a critical value determined based on passenger comfort research; exceeding this threshold significantly deteriorates the passenger experience. In the emergency response plan, temporary shuttle services are prioritized for the most congested sections. Departure intervals are dynamically adjusted based on load factor excesses. Linking with surrounding routes identifies possible alternative routes through road network analysis. The passenger information push system integrates with major map and travel apps to ensure timely information delivery.

[0106] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for monitoring and dispatching public transportation capacity, characterized in that: The method comprises: Step S1, standardizing and detecting anomalies in the collected data of the target bus route to generate preprocessed data, wherein the collected data includes real-time passenger flow, historical passenger flow, and environmental factor data of the target bus route; Step S2: Based on the pre-processed data, the passenger flow of the target bus route in the future period is predicted using the constructed hybrid prediction model MM to obtain the predicted passenger flow. The hybrid prediction model MM includes a time series decomposition model TM and a random fluctuation model RM; the time series decomposition model TM decomposes the passenger flow into basic passenger flow, periodic passenger flow, and trend passenger flow; the random fluctuation model RM captures short-term random fluctuations; The expression of the time series decomposition model TM is: ;in, is the total passenger flow at time t predicted; basic passenger flow Indicates a fixed minimum passenger flow level, which is the average of the lowest passenger flow in the same period of history; Periodic passenger flow Indicates the cyclical passenger flow changes obtained by daily and weekly statistics: ,in and are the daily cycle amplitude and phase, and are the periodic amplitude and phase, respectively; Trending passenger flow Indicates the long-term trend, calculated by the moving average method; Indicates short-term random fluctuations; Step S3, dividing the vehicles to be dispatched into fixed shifts FT, flexible shifts AT and standby shifts BT, and obtaining the predicted full load factor of the target bus route based on the predicted passenger flow; Forecast load factor The calculation formula is: ,in, is the number of running vehicles at time t, The rated passenger capacity of the bicycle; Based on the predicted full load factor, dynamically adjust vehicle configuration using transport scheduling rules; The vehicles to be dispatched are scheduled using 60% fixed shifts FT and 40% flexible shifts AT. The transport capacity dispatching rules include: When the predicted full load factor When the number of mobile shifts AT is reduced, the proportion of reduced mobile shifts AT is , but the maximum reduction rate of mobile shift AT shall not exceed 40%; When the load factor is predicted When maintaining the current capacity configuration; When the load factor is predicted When the mobile shift AT is increased, the proportion of the increased mobile shift AT is ; When the load factor is predicted When the emergency stops, all standby BT shifts will be activated and vehicles on nearby routes will be dispatched.

2. The public transportation capacity monitoring and dispatching method according to claim 1, characterized in that: The random fluctuation model RM is a model established based on environmental factor data, and the environmental factor data includes: weather influencing factors and holiday impact factors , the calculation formula of the stochastic fluctuation model RM is: ,in, is the standard deviation of passenger flow in the same historical period, is a random variable that follows a standard normal distribution.

3. The public transportation capacity monitoring and dispatching method according to claim 2, characterized in that: The hybrid prediction model MM is evaluated by the mean absolute percentage error , root mean square error, and coefficient of determination are evaluated: Mean absolute percentage error: ; Root mean square error ; Coefficient of determination ; when or When the model parameter adjustment mechanism is triggered, the basic passenger flow is recalculated. and periodic passenger flow , update the hybrid prediction model MM.

4. The public transportation capacity monitoring and dispatching method according to claim 3, characterized in that: The predicted passenger flow is obtained by using a multi-time scale sliding window for passenger flow prediction. The multi-time scale sliding window includes: a short-term prediction window, a medium-term prediction window and a long-term prediction window. The short-term prediction window SW is used to predict passenger flow changes in the next 30-60 minutes, with a sampling frequency of 5 minutes; the medium-term prediction window MW is used to predict passenger flow changes in the next 3-6 hours, with a sampling frequency of 15 minutes; the long-term prediction window LW is used to predict passenger flow changes in the next 24 hours, with a sampling frequency of 30 minutes; The short-term prediction window SW relies on the real-time data of the last 30 minutes and a simplified prediction model, with the calculation delay controlled within 10 seconds. The hybrid prediction model expression of the short-term prediction window is: ;in, The original definition of basic passenger flow is retained, indicating a fixed minimum passenger flow level, which is taken as the average of the lowest passenger flow in the same period in history; Considering only the periodic pattern of the last 30 minutes, it simplifies to: ; For real-time random fluctuations, the simplified calculation is: ;in is the standard deviation of passenger flow in the last 30 minutes; The medium-term forecast window MW combines the accumulated data of the day and the historical data of the same period. The hybrid forecast model expression of the medium-term forecast window is: ； in, The original basic passenger flow definition is retained; Consider the cyclical pattern for the day: ; To simplify the trend passenger flow, only the moving average of the last 7 days is considered; For medium-term stochastic fluctuations: ; is the standard deviation of passenger flow during the same period of the day.

5. A public transportation capacity monitoring and dispatching system, used to implement the public transportation capacity monitoring and dispatching method according to any one of claims 1 to 4, characterized in that: The system includes: a data acquisition module, a passenger flow prediction module and a transportation capacity scheduling module; The data acquisition module is used to collect real-time passenger flow, historical passenger flow and environmental factor data of the target bus route; it is used to standardize and detect anomalies in the collected data; the passenger flow prediction module is used to establish a hybrid prediction model MM to predict passenger flow in different time periods in the future; the capacity scheduling module is used to dynamically adjust vehicle configuration according to the predicted passenger flow.

6. The public transportation capacity monitoring and dispatching system according to claim 5, characterized in that: The data acquisition module includes: The on-board passenger flow statistics unit collects passenger flow data on getting on and off the vehicle through infrared sensors or weight sensors; the station passenger flow monitoring unit monitors the number of passengers waiting at the station through cameras and intelligent video analysis systems; the vehicle GPS positioning unit collects real-time vehicle location, speed and operating status information; the environmental information collection unit obtains weather information; all of the above units are connected to the data processing system of the data collection module in real time through a wireless communication network.

7. The public transportation capacity monitoring and dispatching system according to claim 6, characterized in that: The system also includes a visual display module and an emergency processing module; The visual display module is used to display the line operation status, vehicle distribution, passenger flow density heat map and full load rate warning information in real time; When the predicted full load rate exceeds 95%, the emergency handling module automatically triggers the emergency plan, including: temporarily adding shuttle buses, adjusting departure intervals, and linking with surrounding bus routes. At the same time, it pushes congestion reminders and alternative travel suggestions to passengers through the public information platform.

Images