Control methods, devices, electronic equipment and storage media for rail transit

By employing a dual-twin architecture and dynamic game model in the rail transit system, safety strategies under normal operation and extreme weather conditions are collaboratively simulated, resolving the conflict between efficiency and safety in rail transit under extreme weather conditions, and improving the stability and safety of the system.

CN122684508BActive Publication Date: 2026-11-10CHINA RAILWAY LIUYUAN GRP CO LTD
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202611186510.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-08-06
Publication Date
2026-11-10
Estimated Expiration
2046-08-06

AI Technical Summary

Technical Problem

Existing rail transit systems struggle to balance routine operational efficiency and safety risks under extreme weather conditions, and a single twin architecture cannot effectively coordinate conflicting decisions between routine operations and emergency safety.

Method used

A dual-twin architecture is adopted, with normal twins and disaster twins established separately. Through dynamic game theory model, a target scheduling strategy that balances efficiency and security is output.

Benefits of technology

It has achieved a comprehensive improvement in the stability and safety of the rail transit system under extreme weather conditions. Through closed-loop control of pre-warning, in-process handling and post-event recovery, it has improved the initiative of emergency response and the accuracy of decision-making.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122684508B_ABST
    Figure CN122684508B_ABST
Patent Text Reader

Abstract

This disclosure provides a control method, apparatus, electronic device, and storage medium for rail transit. The method includes: acquiring operational data and environmental data of the rail transit; using a first twin, determining a first set of strategies that meet the operational efficiency requirements of the rail transit based on the operational data and the environmental data; wherein the first set of strategies is a set of scheduling strategies for the rail transit under normal conditions; using a second twin, determining a second set of strategies that meet the safety risks of the rail transit based on the operational data and the environmental data; wherein the second set of strategies is a set of scheduling strategies for the rail transit under disaster conditions; using a dynamic game model, determining a target scheduling strategy for the rail transit based on the first set of strategies and the second set of strategies, and controlling the rail transit based on the target scheduling strategy. This achieves a target scheduling strategy that balances efficiency and safety.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This disclosure relates to the field of data processing technology, and in particular to a control method, device, electronic equipment and storage medium for rail transit. Background Technology

[0002] As the coverage of rail transit continues to expand, the complexity of daily operations and the pressure on emergency safety under extreme weather conditions are increasing simultaneously. However, using a single twin to determine the scheduling strategy for rail transit may result in a failure to simultaneously address both routine operations and safety risks.

[0003] In view of this, how to simultaneously address routine operations and security risks has become an urgent technical problem to be solved. Summary of the Invention

[0004] In view of this, the purpose of this disclosure is to provide a control method, device, electronic equipment and storage medium for rail transit to solve or partially solve the above-mentioned technical problems.

[0005] To achieve the above objectives, the first aspect of this disclosure proposes a control method for rail transit, the method comprising: Acquire operational and environmental data of rail transit; Using a first twin, a first set of strategies that meet the operational efficiency of rail transit is determined based on the operational data and the environmental data; wherein, the first set of strategies is the set of scheduling strategies when the rail transit is in a normal state; Using a second twin, a second set of strategies that meet the safety risks of rail transit is determined based on the operational data and the environmental data; wherein, the second set of strategies is a set of scheduling strategies for rail transit under disaster conditions; Using a dynamic game model, a target scheduling strategy for the rail transit is determined based on the first strategy set and the second strategy set, and the rail transit is controlled based on the target scheduling strategy.

[0006] Based on the same inventive concept, a second aspect of this disclosure proposes a control device for rail transit, comprising: The acquisition module is configured to acquire operational and environmental data of the rail transit system. The first strategy set determination module is configured to use a first twin to determine a first strategy set that satisfies the operational efficiency of rail transit based on the operational data and the environmental data; wherein, the first strategy set is the scheduling strategy set of the rail transit under normal conditions; The second strategy set determination module is configured to use a second twin to determine a second strategy set that meets the rail transit safety risk based on the operational data and the environmental data; wherein, the second strategy set is the scheduling strategy set of the rail transit under disaster conditions; The target scheduling strategy determination module is configured to use a dynamic game model to determine the target scheduling strategy of the rail transit based on the first strategy set and the second strategy set, and to control the rail transit based on the target scheduling strategy.

[0007] Based on the same inventive concept, a third aspect of this disclosure proposes an electronic device including a memory, a processor, and a computer program stored in the memory and executable by the processor, wherein the processor implements the method described above when executing the computer program.

[0008] Based on the same inventive concept, a fourth aspect of this disclosure provides a non-transitory computer-readable storage medium that stores computer instructions for causing a computer to perform the methods described above.

[0009] As can be seen from the above, the control method, device, electronic equipment, and storage medium for rail transit provided in this disclosure acquire rail transit operation data and environmental data. Using a first twin, a first set of strategies that meets the operational efficiency of the rail transit is determined based on the operational and environmental data; wherein, the first set of strategies is the set of scheduling strategies for the rail transit under normal conditions. Using a second twin, a second set of strategies that meets the safety risks of the rail transit is determined based on the operational and environmental data; wherein, the second set of strategies is the set of scheduling strategies for the rail transit under disaster conditions. Thus, by using the first and second twins, respectively outputting a first set of strategies adapted to normal operational efficiency and a second set of strategies adapted to extreme weather safety risks, the problem that a single twin architecture cannot simultaneously address both normal operation and safety risks can be solved. Using a dynamic game model, a target scheduling strategy for the rail transit is determined based on the first and second strategy sets, and the rail transit is controlled based on the target scheduling strategy. Thus, by using a dynamic game model to collaboratively deduce the first and second strategy sets, a target scheduling strategy that balances efficiency and safety can be accurately output, thereby comprehensively improving the stability and safety of rail transit operation. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in this disclosure or related technologies, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a flowchart of a rail transit control method according to an embodiment of the present disclosure; Figure 2 This is a flowchart of a digital twin-based rail transit control system according to an embodiment of the present disclosure; Figure 3 This is a timing diagram of the hierarchical early warning switching in an embodiment of this disclosure; Figure 4 This is a schematic diagram of the solution space of the dynamic game model in an embodiment of this disclosure; Figure 5 This is a flowchart of a digital twin-based rail transit control method according to an embodiment of the present disclosure; Figure 6 This is a schematic diagram comparing the implementation effects of the embodiments of this disclosure; Figure 7 This is a schematic diagram of the structure of a rail transit control device according to an embodiment of the present disclosure; Figure 8 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present disclosure. Detailed Implementation

[0012] To make the objectives, technical solutions, and advantages of this disclosure clearer, the following detailed description is provided in conjunction with specific embodiments and the accompanying drawings.

[0013] It should be noted that, unless otherwise defined, the technical or scientific terms used in the embodiments of this disclosure should have the ordinary meaning understood by one of ordinary skill in the art to which this disclosure pertains. The terms "first," "second," and similar terms used in the embodiments of this disclosure do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0014] Based on the background description, urban rail transit undertakes the task of large-scale passenger flow distribution in cities. The operational safety, punctuality, and service continuity of rail transit are directly related to public safety, urban operational efficiency, and social stability. In recent years, extreme weather events have shown a trend of increasing frequency, intensity, scope of impact, and suddenness. Typhoons may cause crosswinds exceeding limits in elevated sections, pantograph disconnection, and decreased lateral stability of trains; rainstorms and floods may cause water accumulation in underground sections, backflow at entrances and exits, pump room overload, and short circuits in power supply equipment; freeze-thaw cycles and low temperatures may cause sudden changes in track structural stiffness, icing of the overhead contact line, and abnormal switch operation; high temperatures may lead to rail expansion, abnormal equipment heat dissipation, and increased risk of heat stress for passengers. These disasters not only affect individual sections but may also have a chain reaction across the entire network through train schedules, passenger transfer relationships, and emergency resource allocation.

[0015] Some solutions utilize digital twins for rail transit emergency management and risk control systems. For example, a digital twin-based emergency command method and system for railway integrated hubs uses a twin model to dynamically assess hazard risks and output a ranking of emergency response plans; an intelligent auxiliary decision-making method for urban rail transit emergency command constructs a full-element digital twin covering trains, tracks, stations, power supply, and emergency resources, and generates emergency dispatch instructions based on the twin's ability to capture abnormal trigger signals; a digital twin-based intelligent flood control system for subways uses sensing equipment and meteorological data for risk warnings and emergency plan recommendations. These solutions improve the informatization and visualization level of rail transit emergency management, but the following technical challenges remain: First, digital twin systems often employ a single twin architecture, making it difficult to simultaneously address the inherently conflicting goals of efficient routine operation and emergency safety during extreme weather. If the weight of safety indicators is increased in the objective function of a single twin, the system is prone to becoming overly conservative under minor weather disturbances, leading to excessive speed limits, shutdowns, or resource pre-provisioning, thus impairing daily operational efficiency. Conversely, if the safety weight is reduced to maintain operational efficiency, insufficient emergency response may occur when extreme weather conditions rapidly develop.

[0016] Second, there is a lack of systematic mechanisms to resolve the decision-making conflicts between routine operations and emergency response. Routine rail transit operations typically aim to minimize delays, maximize capacity, reduce energy consumption, and maintain passenger service quality; while emergency response to extreme weather conditions aims to minimize accident risk, maximize safety margins, and prioritize evacuation reliability and system resilience. These two approaches directly conflict on key actions such as speed limits, service suspensions, turnarounds, closures, passenger evacuation, resource allocation, and resumption of operations. Recommendations are often based on empirical rules or fixed weights, making it difficult to quantify whether speed reductions are necessary, to what extent, when to suspend operations, when to resume operations, and to what operational level.

[0017] Third, most solutions involve multi-algorithm collaboration within the same task scenario. For example, multiple twins are created for the same physical entity or the same prediction task, and then the prediction results are merged to improve accuracy. This addresses the problem of "how to integrate different algorithm models under the same objective task." The embodiments disclosed herein differ from these solutions. They address two different objective tasks: routine operation and extreme weather emergency response. Complete and independently operating routine twins and disaster twins are created separately, and a dynamic game theory model is used to coordinate the strategic conflicts between the two. This solves the problem of how to make multi-stage equilibrium decisions under conflicting objective tasks.

[0018] Fourth, emergency dispatch methods tend to focus on post-event response, lacking an integrated closed loop of proactive prevention, ongoing simulations, and smooth post-event recovery. In actual operation, extreme weather risks often evolve gradually from concern, warning, alert to crisis, and dispatch decisions should be dynamically adjusted according to changes in risk. Some plans can provide emergency solutions after a risk is triggered, but there is a lack of complete mechanisms for proactive pre-emptive measures before the risk fully materializes, phased recovery after the risk decreases, and parameter learning after reviewing the execution results.

[0019] Therefore, there is a need for an integrated method and system for early warning and emergency response that can simultaneously maintain the ability to assess normal operational efficiency and the ability to predict disaster safety on the same rail transit physical system, and can quantitatively solve the conflict between the two through game theory.

[0020] As mentioned above, how to balance routine operations and security risks has become an important research question.

[0021] Based on the above description, such as Figure 1 As shown in this embodiment, the control method for rail transit includes: Step 101: Obtain the operation data and environmental data of the rail transit system.

[0022] In practice, operational and environmental data of the rail transit system are acquired and integrated into a multi-source sensing data set. , in, For the current stage A collection of multi-source sensing data. For meteorological data, For infrastructure monitoring data, For operational data, For environmental data, For emergency resource data.

[0023] The aforementioned multi-source data sets, after being processed using a unified spatiotemporal reference, are simultaneously input into both the first and second twin datasets. The unified spatiotemporal reference is established based on the BeiDou Navigation Satellite System (BDS) or the Global Positioning System (GPS), rail transit mileage markers, station numbers, section numbers, and a unified timestamp. For any given sensing data point... Its spatiotemporal label is represented as: , in, For the first Spatiotemporal labels for sensory data Longitude coordinates Latitude coordinates For line numbering, For mileage markers or interval numbers, For timestamps, This serves as a data quality identifier. Spatiotemporal tags ensure that the understanding of the same physical state by the first and second twins is comparable.

[0024] The first and second twins share multi-source sensing data from vehicles, tracks, stations, power supply, signals, environment, meteorology, and emergency resources, but they are independent of each other in terms of model composition, state variables, objective functions, inference frequency, and dominant conditions.

[0025] Step 102: Using the first twin, determine a first set of strategies that meet the operational efficiency of rail transit based on the operational data and the environmental data; wherein, the first set of strategies is a set of scheduling strategies for rail transit under normal conditions.

[0026] In practice, the first twin is deployed in a central twin server cluster in the cloud center, storing the full-element Building Information Model (BIM), route topology model, vehicle dynamics model, timetable model, passenger flow forecast model, energy consumption model, and equipment health status model. The first twin uses train punctuality rate, section capacity, passenger service quality, and energy efficiency as optimization objectives for daily status monitoring, timetable adjustments, capacity matching, energy consumption optimization, and efficiency assessment.

[0027] The first twin can be a normal twin. The first strategy set is the set of candidate scheduling strategies output by the normal twin based on the current operational status.

[0028] At the current stage The first state vector of the first twin is represented as: , in, For the set of states of the first twin, For train location gathering, For the set of train speeds, This refers to the train tracking interval status. In energy consumption state, For station and section passenger flow status, For the health status of critical equipment, This indicates a deviation from the running graph.

[0029] Step 103: Using the second twin, determine a second set of strategies that meet the safety risks of rail transit based on the operational data and the environmental data; wherein, the second set of strategies is a set of scheduling strategies for rail transit under disaster conditions.

[0030] In practical implementation, the second twin is deployed on edge twin server clusters or high-availability computing nodes in the dispatch center, storing extreme weather evolution models, line vulnerability models, facility failure probability models, passenger evacuation models, emergency resource scheduling models, and system resilience recovery models. The second twin uses accident risk probability, passenger safety threats, system resilience loss, resource arrival time, and recovery time as optimization objectives for rapid extreme weather simulation, risk trend prediction, emergency alternative solution generation, and recovery solution evaluation.

[0031] The second twin can be a disaster twin. The second strategy set is the set of emergency response strategies on the safety side output by the second twin based on the risk simulation results.

[0032] At the current stage The second state vector of the second twin is represented as: , in, For the set of states of the second twin, This is the meteorological input status. As a disaster risk field, The facility is in a vulnerable state. The scope of the disaster's impact. Passenger evacuation status In emergency resource status, This represents the system's resilience state.

[0033] Step 104: Using a dynamic game model, determine the target scheduling strategy for the rail transit based on the first strategy set and the second strategy set, and control the rail transit based on the target scheduling strategy.

[0034] In specific implementation, this embodiment of the disclosure sets the first twin and the second twin as two game modules. The first twin tends to maintain traffic efficiency, reduce delays, reduce passenger backlog, and control energy consumption, while the second twin tends to reduce the probability of accidents, increase safety margins, shorten evacuation time, and ensure system resilience. The first and second twins output corresponding scheduling strategies at each time stage, and the dynamic game model solves for the equilibrium scheduling strategy (i.e., the target scheduling strategy) within multiple stage prediction windows.

[0035] The dynamic game model employs a rolling time-domain approach for solving the problem. For example, if each stage is 10 minutes long and the model predicts the next 2 hours forward, each solution involves 12 stages. The system only executes the optimal action for the current stage and re-solves the problem after receiving new sensor data in the next stage, thus adapting to rapid changes in extreme weather.

[0036] The target dispatching strategy includes: train speed limits, service suspensions, detours, or turnaround instructions; time windows for section closures and reopenings; emergency resource allocation routes; station evacuation and passenger flow control plans; passenger information dissemination content; and a plan for restoring normal operations. The restoration plan includes at least one of the following: restoration phase divisions, restoration conditions for each phase, estimated restoration time, risk review indicators, and fallback actions in case of restoration failure.

[0037] The target scheduling strategy output by the dynamic game model, after instruction conversion, first enters the safety constraint verification and scheduling confirmation process before being sent to the rail transit operation control system. Train scheduling suggestions first enter the safety constraint verification module to verify whether the instructions meet the constraints of block, interlocking, speed limits, platform capacity, and timetable. After passing the verification, they are confirmed by the dispatcher or existing train operation safety rules, and then sent to the automatic train operation system and automatic train monitoring system via standardized interfaces. Passenger evacuation guidance plans are pushed through the passenger information system, broadcast system, station display screens, and mobile devices; emergency resource allocation routes are pushed through the emergency command system to on-site rescue teams, inspection personnel, drainage and emergency repair personnel, and material support personnel.

[0038] Figure 2 This is a flowchart of a digital twin-based rail transit control system according to an embodiment of this disclosure. Figure 2As shown, the rail transit control system based on a digital twin includes: a multi-source sensing data acquisition and processing module, a normal state twin, a disaster twin, a dynamic game decision-making module, command execution and closed-loop feedback, and a rail transit operation control system. The multi-source sensing data acquisition and processing module (weather station, sensors, cameras, radar, on-board equipment) acquires data on extreme weather environments / the physical world of rail transit. The normal state twin is deployed on a cloud-based server cluster and includes a BIM model, a vehicle dynamics model, and an operation scheduling model. The disaster twin is deployed on an edge server cluster and includes an extreme weather evolution model, an emergency evacuation module, and equipment vulnerability assessment. The dynamic game decision-making module includes a Nash equilibrium solver.

[0039] This disclosure pertains to the fields of intelligent operation and maintenance, safety emergency response, digital twins, disaster early warning, and intelligent dispatching decision-making in the rail transit industry. Specifically, it relates to an integrated method and system for early warning and emergency response in urban rail transit, intercity railways, suburban railways, and railway hubs under the influence of extreme weather events such as typhoons, rainstorms, floods, freeze-thaw cycles, high temperatures, and strong winds. This disclosure establishes a first twin for optimizing routine operational efficiency and a second twin for optimizing safety resilience in extreme weather. A dynamic game model is used to perform multi-stage trade-offs on the strategy sets output by these two twins, achieving closed-loop control of the rail transit system from pre-event early warning, in-event response, post-event recovery to strategy self-learning.

[0040] The rail transit system in this disclosure includes at least one of the following: underground lines, elevated lines, ground lines, depots, parking lots, transfer hubs, power supply systems, communication and signaling systems, water supply and drainage systems, passenger service systems, and emergency resource dispatch systems. This disclosure is particularly applicable to scenarios where there is a significant conflict between routine operational efficiency goals and emergency safety goals in extreme weather conditions. Examples include whether to maintain low-speed operation on elevated sections under strong wind conditions, whether to close underground sections under heavy rain conditions, whether to advance maintenance windows under freeze-thaw conditions, and whether to adjust train intervals under extreme high-temperature conditions.

[0041] The embodiments disclosed herein aim to address the technical problems of digital twin technology for rail transit in extreme weather emergency scenarios, such as the inability of a single twin to simultaneously balance normal operation efficiency and emergency safety margin, the lack of multi-stage quantitative trade-offs in emergency dispatch decisions, the inapplicability of multi-twin solutions to conflicting coordination of different objectives and tasks, and the lack of proactive early warning and mode transfer mechanisms in emergency systems.

[0042] The purpose of this disclosure is to provide an integrated method and system for early warning and emergency response to extreme weather in rail transit based on dual digital twins and dynamic game theory. By establishing a dual-twin architecture with a first twin and a second twin operating in parallel, the first twin continuously evaluates train operation efficiency, passenger service quality, energy consumption, and timetable stability, while the second twin continuously or according to early warning levels evaluates disaster evolution, facility vulnerability, passenger evacuation risks, and system resilience. Furthermore, a dynamic game model is used to solve for equilibrium strategy combinations that satisfy hard safety constraints at multiple time stages, achieving a computable, interpretable, and iterative equilibrium between normal operation and extreme weather emergency response.

[0043] The embodiments disclosed herein also aim to provide a closed-loop mechanism from early warning triggering, disaster simulation, alternative solution generation, game decision-making, instruction execution, mode transfer to strategy self-learning, so as to transform the safety management of rail transit in extreme weather from a reactive response to an active defense throughout the entire process.

[0044] Through the above embodiments, operational and environmental data of the rail transit are acquired. Using a first twin, a first set of strategies that meets the operational efficiency requirements of the rail transit is determined based on the operational and environmental data; wherein, the first set of strategies is the set of scheduling strategies for the rail transit under normal conditions. Using a second twin, a second set of strategies that meets the safety risks of the rail transit is determined based on the operational and environmental data; wherein, the second set of strategies is the set of scheduling strategies for the rail transit under disaster conditions. Thus, by using the first and second twins, respectively outputting a first set of strategies adapted to normal operational efficiency and a second set of strategies adapted to extreme weather safety risks, the problem that a single twin architecture cannot simultaneously address both normal operation and safety risks can be solved. Using a dynamic game theory model, a target scheduling strategy for the rail transit is determined based on the first and second strategy sets, and the rail transit is controlled based on the target scheduling strategy. Thus, by using a dynamic game theory model to collaboratively deduce the first and second strategy sets, a target scheduling strategy that balances efficiency and safety can be accurately output, thereby comprehensively improving the stability and safety of rail transit operation.

[0045] In some embodiments, step 103 includes: Step 103A: Obtain real-time meteorological values ​​and safety reference values ​​for various meteorological indicators, and determine the meteorological hazard index based on the real-time meteorological values ​​and the safety reference values.

[0046] In practice, meteorological indicators include at least one of the following: rainfall intensity, wind speed, temperature, snow depth, and lightning intensity.

[0047] The meteorological hazard index is determined based on real-time meteorological values ​​and safety reference values. , in, For the current stage The meteorological hazard index, The total number of meteorological indicators. For the first Real-time or forecasted meteorological values ​​of meteorological indicators. For the first Safety reference values ​​for meteorological indicators For the first Danger thresholds for meteorological indicators For the first Weights of meteorological indicators.

[0048] In the above formula, the safety reference value and danger threshold It is determined by the route operation rules, meteorological disaster levels, and historical events.

[0049] Step 103B: Obtain the risk coefficients of multiple routes, and perform weighted processing on the risk coefficients of the multiple routes to obtain the route hazard index.

[0050] In practice, the risk index of a railway line is related to its spatial environment, engineering structure, and historical disaster records. The risk coefficients for various types of lines include: underground lines, elevated lines, lines near water or in low-lying areas, geological conditions, and sections with steep curves.

[0051] The risk coefficients of various routes are weighted to obtain the route hazard index. , in, For the current stage The risk level of the route. This is the first weighting coefficient corresponding to the risk coefficient of underground lines. For the risk coefficient of underground lines, This is the second weighting coefficient corresponding to the risk coefficient of the elevated line. The risk factor for elevated lines, This is the third weighting coefficient corresponding to the risk coefficient of low-lying areas near water. The risk factor for low-lying areas near water. This is the fourth weighting coefficient corresponding to the geological condition risk coefficient. Geological condition risk coefficient, This is the fifth weighting coefficient corresponding to the risk coefficient of the curve slope section. The risk coefficient for sections with small curve radii or steep slopes.

[0052] Step 103C: Obtain real-time monitoring values ​​and historical normal averages of various monitoring indicators, and determine the indicator abnormality index based on the real-time monitoring values ​​and the historical normal averages.

[0053] In practice, the anomaly index is used to reflect the degree of deviation between the real-time monitoring values ​​collected by sensors and the historical average values ​​under normal conditions. The monitoring indicators include at least one of the following: track, overhead contact line, tunnel, water level, wind speed, temperature, and signal equipment.

[0054] The abnormal index is determined based on real-time monitoring values ​​and historical normal averages. , in, For the current stage Abnormal index of indicators The total number of monitoring indicators, For the first Real-time monitoring values ​​of each monitoring indicator For the first Historical normal average values ​​of each monitoring indicator For the first The historical standard deviation of each monitoring indicator To prevent tiny positive numbers with a denominator of zero.

[0055] In the above formula, the historical normal mean and historical standard deviation It is determined by historical data from the same route, the same season, and similar operating conditions.

[0056] Step 103D: Obtain the frequency parameters, impact parameters, average recovery time, and early warning error correction value of historical disasters; and perform weighted processing on the frequency parameters, impact parameters, average recovery time, and early warning error correction value to obtain the historical disaster index.

[0057] In practice, the frequency of disasters within the historical statistical period is obtained, and the frequency parameters of historical disasters are determined based on the disaster frequency. , in, For the current stage The frequency of historical disasters This refers to the number of times the same or similar disasters occurred within the target area during the historical statistical period and reached the operational impact threshold. For reference frequency, It takes a smaller value to limit the normalization result to the range of 0 to 1.

[0058] The historical disaster index is used to reflect the sensitivity of the same line, the same section, or similar engineering environment to operational interruptions, equipment malfunctions, water accumulation, crosswinds exceeding limits, abnormal freeze-thaw cycles, or passenger evacuation pressures under historical extreme weather conditions.

[0059] Frequency parameters of historical disasters Determined jointly by line grade, operational intensity, and historical statistical period. Parameters regarding the degree of impact caused by historical disasters. The average recovery time for historical disasters can be calculated by weighting historical maximum speed limit duration, downtime, number of turnarounds, number of passenger clearings, and equipment failure level. This is obtained by dividing the historical recovery time by the reference recovery time. The early warning error correction value for historical disasters. It is used to reflect the situation of underreporting, late reporting, or underestimation of risks in line sections during historical statistical periods.

[0060] The frequency parameter, impact parameter, mean recovery time, and early warning error correction value are normalized separately to avoid inconsistencies in the dimensions of the parameters. The normalized frequency parameter, impact parameter, mean recovery time, and early warning error correction value are then weighted to obtain the historical disaster index. , in, For the current stage Historical disaster index, This is the first weighting coefficient corresponding to the frequency parameter. For the current stage Frequency parameters of historical disasters This is the second weighting coefficient corresponding to the degree of influence parameter. For the current stage Parameters on the impact of historical disasters The third weighting coefficient corresponding to the average recovery time. For the current stage Average recovery time for historical disasters This is the fourth weighting coefficient corresponding to the early warning error correction value. For the current stage Error correction values ​​for early warning of historical disasters.

[0061] In the above formula, the sum of the multiple weight coefficients is 1, meaning that the weight coefficients satisfy the following condition: .

[0062] Step 103E: The meteorological hazard index, the line hazard index, the indicator anomaly index, and the historical disaster index are weighted to obtain a comprehensive early warning index.

[0063] In practice, the meteorological hazard index, line hazard index, indicator anomaly index, and historical disaster index are weighted to obtain a comprehensive early warning index. , in, For the current stage The comprehensive early warning index, This is the first weighting coefficient corresponding to the meteorological hazard index. For the current stage The meteorological hazard index, This is the second weighting coefficient corresponding to the line hazard index. For the current stage The risk level of the route. This is the third weighting coefficient corresponding to the abnormal index. For the current stage Abnormal index of indicators This is the fourth weighting coefficient corresponding to the historical disaster index. For the current stage Historical disaster index.

[0064] In the above formula, the sum of the multiple weight coefficients is 1, meaning that the weight coefficients satisfy the following condition: .

[0065] Step 103F: Determine the warning level based on the comprehensive warning index, and determine the operating frequency based on the warning level.

[0066] In practice, the warning level is determined based on the comprehensive warning index, and the operating frequency is determined based on the warning level. The warning levels include at least one of the following: blue warning, yellow warning, orange warning, and red warning.

[0067] Different lines can have their early warning index thresholds differentiated based on factors such as the proportion of underground lines, the proportion of elevated lines, geological conditions, surrounding water systems, station entrance and exit elevations, pumping station capacity, and the frequency of historical disasters. When the early warning level rises, the second twin increases the simulation frequency and expands the prediction window; when the early warning level falls, the second twin gradually reduces the simulation frequency, but continues to monitor key safety indicators throughout the relocation period to avoid premature recovery and secondary risks.

[0068] Step 103G: Invoke the second twin according to the operating frequency, and use the second twin to determine the second strategy set based on the operating data and the environmental data.

[0069] In specific implementation, as shown in Table 1, when the comprehensive early warning index... Within the warning index range Within this system, the warning level is set to blue. The second twin generates a second strategy set at a frequency of once per hour. When the comprehensive warning index... Within the warning index range Within this context, the warning level is determined to be yellow. The second twin generates a second strategy set at a frequency of once every 10 minutes. When the comprehensive warning index... Within the warning index range Within this system, the warning level is set to orange. The second twin generates a second strategy set at a frequency of once per minute. When the comprehensive warning index... When the warning level is determined to be red, the second twin continuously solves the second strategy set.

[0070] Table 1. Mapping Relationship between Warning Levels and Operation Frequency

[0071] Figure 3 This is a timing diagram of the hierarchical early warning switching in an embodiment of this disclosure. Figure 3 As shown, in the first stage t0, the warning level is zero, the normal twin is in control, the disaster twin is on standby, dynamic game decision-making is not activated, and the execution system is normal. In the second stage t1, the warning level is blue, the normal twin is in control, the disaster twin scans, dynamic game decision-making is rehearsed, and the execution system issues a warning. In the third stage t2, the warning level is yellow, the normal twin is on standby, the disaster twin is in real-time, dynamic game decision-making is low-frequency, and the execution system is preset. In the fourth stage t3, the warning level is orange, the normal twin is backed up, the disaster twin is at full speed, dynamic game decision-making is high-frequency, and the execution system partially executes. In the fifth stage t4, the warning level is red, the normal twin is backed up, the disaster twin is at full speed, dynamic game decision-making is forced, and the execution system fully executes. In the sixth stage t5, the warning level recovers from red to green, the normal twin resumes control, the disaster twin reduces its frequency, the dynamic game decision-making strategy is reverted, and the execution system reverts to its strategy.

[0072] The above scheme obtains real-time meteorological values ​​and safety reference values ​​for various meteorological indicators, and determines the meteorological hazard index based on these values. Risk coefficients for various lines are obtained, and a weighted average is applied to obtain the line hazard index. Real-time monitoring values ​​and historical normal averages for various monitoring indicators are obtained, and an indicator anomaly index is determined based on these values. Frequency parameters, impact parameters, average recovery time, and warning error correction values ​​for historical disasters are obtained, and a weighted average is applied to obtain the historical disaster index. A comprehensive warning index is obtained by weighting the meteorological hazard index, line hazard index, indicator anomaly index, and historical disaster index. The warning level is determined based on the comprehensive warning index, and the operating frequency is determined based on the warning level. A second twin is invoked based on the operating frequency, and the second twin is used to determine a second strategy set based on operating data and environmental data. Thus, by establishing a tiered warning triggering mechanism, the activation level, operating frequency, and whether the second twin has entered a game-theoretic decision-making state can be determined. This embodiment does not directly adopt the meteorological department's warning levels, but instead integrates the meteorological hazard index, the line hazard index, the indicator anomaly index, and the historical disaster index to form a comprehensive warning index applicable to rail transit.

[0073] In some embodiments, step 103 includes: Step 1031: Obtain current weather data and a rail transit route topology map, and determine and update the disaster risk field based on the current weather data and the route topology map.

[0074] In practice, after receiving early warning information, the second twin dynamically extrapolates the development trend of the disaster risk field based on real-time sensing data and meteorological forecast data. Specifically, it determines and updates the disaster risk field based on current weather data and route topology maps. , in, For the next stage Updated disaster risk field Let be the risk field evolution function. For the current stage The current disaster risk field, For the current stage Current weather data, This is a diagram of the route topology. For time step, This represents the model perturbation term.

[0075] Among them, the risk field can be a wind-induced risk field, a water accumulation risk field, a freeze-thaw risk field, or a multi-hazard coupled risk field.

[0076] Step 1032: Obtain weather intensity index, disaster environment index, train operating speed, section hazard coefficient and passenger parameters, and determine the risk probability of rail transit based on the weather intensity index, the disaster environment index, the train operating speed, the section hazard coefficient and the passenger parameters.

[0077] In specific implementation, for line sections The second twin determines the risk probability of rail transit based on weather intensity indicators, disaster environment indicators, train operating speed, section hazard coefficient, and passenger parameters. , in, For line section At the current stage The probability of risk, For the Sigmoid function, For constant terms, As the first parameter, For line section At the current stage Weather intensity index For the second parameter, For line section At the current stage Disaster environmental indicators As the third parameter, For line section At the current stage The train's operating speed, The fourth parameter, For line section The section risk factor, The fifth parameter, For line section At the current stage Passenger parameters.

[0078] In the above formula, the weather intensity index These can be wind speed, rainfall intensity, or temperature. Hazardous environmental indicators. This could be water level, snow depth, or ice thickness. Passenger parameters. This could be passenger density or passenger exposure. (First parameter) Second parameter Third parameter The fourth parameter and the fifth parameter These can be regression coefficients or model parameters.

[0079] Step 1033: Obtain the first functional parameters of the rail transit under disaster conditions and the second functional parameters under normal conditions, and determine the system loss of the rail transit based on the first functional parameters and the second functional parameters.

[0080] In practice, the system loss of rail transit is determined based on the first functional parameter of rail transit under disaster conditions and the second functional parameter under normal conditions. , in, For the current stage The system loss, For the current stage The primary functional parameter of rail transit under disaster conditions. For the current stage The second functional parameter of rail transit under normal conditions.

[0081] In the above formula, the system loss It could be a system resilience loss, the first functional parameter. It can be the remaining service capacity of the rail transit system under disaster conditions, a second functional parameter. This can refer to the overall service capacity of a rail transit system under normal conditions. (First functional parameter) It can be determined comprehensively by the proportion of operational sections, the number of available trains, the number of available stations, the number of available transfer channels, and the proportion of available emergency resources. When the system suffers a loss... The larger the value, the more severe the damage to the rail transit system's line function under disaster conditions.

[0082] Step 1034: Generate a scheduling strategy template based on the updated disaster risk field, the risk probability, and the system loss, and generate candidate scheduling strategies based on the scheduling strategy template.

[0083] In practice, the second twin generates a scheduling strategy template based on updated disaster risk fields, risk probabilities, and system losses, and then generates candidate scheduling strategies based on the scheduling strategy template. The candidate scheduling strategy includes at least one emergency response alternative. Each alternative can be represented as: ,in, For the first One emergency response alternative plan For train dispatching actions, For station control actions, Actions to guide passenger evacuation. For emergency resource allocation actions, Actions to resume operations.

[0084] The alternative plan includes at least one of the following: speed-limited or suspended sections, train turnaround locations, deployment time of backup trains, evacuation routes to affected stations, station passenger flow control plans, dispatch routes for emergency supplies and rescue personnel, emergency operation modes for power supply and drainage systems, and the division of normal operation recovery phases.

[0085] Step 1035: Retrieve the set of security constraints and determine the second scheduling strategy based on the candidate scheduling strategy and the set of security constraints.

[0086] In practice, the set of safety constraints includes multiple safety constraints, and candidate scheduling strategies must simultaneously satisfy multiple safety constraints. These safety constraints include: block safety constraints, interlocking safety constraints, platform capacity constraints, passenger evacuation capacity constraints, emergency resource arrival time constraints, and maximum permissible risk constraints. The set of safety constraints is represented as follows: , in, For the current stage The set of safety constraints. For blockage safety constraints, For interlocking safety constraints, Due to platform capacity constraints, Constrained by passenger evacuation capacity. Due to constraints on the arrival time and quantity of emergency resources, This is the maximum permissible risk constraint.

[0087] When the alternative schemes in the candidate scheduling strategy If any security constraint is not met, the policy is removed from the candidate scheduling policy to obtain the second policy set. , in, For the current stage The second set of strategies after filtering by security constraints. The first in the candidate scheduling strategy One alternative plan As candidate scheduling strategies, As an alternative plan In state Next to the The degree of violation of class-specific safety constraints, For the current stage The set of safety constraints.

[0088] Second strategy set The scheduling scheme in the second strategy set can be incorporated into the subsequent dynamic game model. If the value is empty, the system will trigger the manual contingency plan rollback mechanism, and the dispatcher will handle the situation according to the existing emergency plan.

[0089] The above scheme obtains current weather data and the rail transit line topology map, and determines an updated disaster risk field based on the current weather data and line topology map. It acquires weather intensity indicators, disaster environment indicators, train operating speed, section hazard coefficients, and passenger parameters, and determines the risk probability of rail transit based on these parameters. It acquires the first functional parameters of rail transit under disaster conditions and the second functional parameters under normal conditions, and determines the system loss of rail transit based on the first and second functional parameters. A scheduling strategy template is generated based on the updated disaster risk field, risk probability, and system loss, and candidate scheduling strategies are generated based on the scheduling strategy template. In this way, the candidate scheduling strategies comprehensively consider the updated disaster risk field, risk probability, and system loss, enabling them to adapt to extreme weather safety risks. A set of safety constraints is retrieved, and a second strategy set is determined based on the candidate scheduling strategies and the set of safety constraints. In this way, the second strategy set can simultaneously satisfy both the candidate scheduling strategies and the set of safety constraints.

[0090] In some embodiments, step 1034 includes: Step 1034A: Generate a scheduling strategy object based on the updated disaster risk field, generate a scheduling strategy level based on the risk probability, and generate a scheduling strategy range based on the system loss.

[0091] In practical implementation, to avoid the second twin being limited to risk assessment, this disclosure further defines an automatic generation mechanism for emergency response alternatives. The alternatives are not arbitrarily generated by the model, but are jointly constrained by scheduling strategy templates, line topology constraints, traffic safety constraints, station capacity constraints, and emergency resource constraints.

[0092] Specifically, the second twin generates scheduling strategy objects based on the updated disaster risk field, generates scheduling strategy levels based on risk probabilities, generates scheduling strategy ranges based on system losses, generates candidate action combinations based on scheduling strategy templates, and finally forms an emergency response alternative plan.

[0093] First, the updated disaster risk field is projected onto the line topology map to identify the affected candidate sections, stations, turnaround nodes, power supply zones, and emergency resource nodes for dispatching strategies. Second, dispatching strategies are categorized into monitoring, speed limits, turnaround, closure, evacuation, and shutdown levels based on the risk probability of each section. Third, the scope of dispatching strategies is determined based on system losses to determine whether it needs to be expanded from a single section to adjacent sections, transfer stations, or cascaded network impact areas. Finally, dispatching strategy templates are invoked to generate candidate action combinations.

[0094] Step 1034B: Generate a scheduling policy template based on the scheduling policy object, the scheduling policy level, and the scheduling policy range.

[0095] In practice, dispatch strategy templates are invoked based on the warning level and disaster type. Dispatch strategy templates include at least one of the following: typhoon and gale template, rainstorm and flooding template, freeze-thaw and low-temperature template, high-temperature track expansion template, and combined disaster template. Each dispatch strategy template provides enumerable candidate action categories, including at least one of the following: speed limit level, section closure range, turnaround station selection, passenger clearing station selection, emergency resource pre-positioning point, and recovery phase.

[0096] Step 1034C: Determine candidate segments from rail transit based on the risk probability, determine candidate actions for the candidate segments from the scheduling strategy template, and combine the candidate actions to obtain a candidate scheduling strategy.

[0097] In practice, based on the rail transit line sections output by the second twin, The risk probability is used to screen candidate segments. For any line segment, when the risk probability... Exceeding the risk threshold Then, add the line segment to the candidate segment set. , in, For the candidate segment set, Represents the first segment in the candidate segment set. Each line section, For line section The probability of risk, This is the risk threshold.

[0098] The scheduling strategy template includes: a set of speed limit levels, a set of blocking actions, a set of turnaround stations, and a set of resource allocation. For each candidate segment in the candidate segment set, candidate speed limit levels are determined from the speed limit level set, candidate blocking actions are determined from the blocking action set, candidate turnaround stations are determined from the turnaround station set, and candidate resource allocations are determined from the resource allocation set. Candidate scheduling strategies are obtained by combining and enumerating the candidate speed limit levels, candidate blocking actions, candidate turnaround stations, and candidate resource allocations.

[0099] , in, As candidate scheduling strategies, For a set of speed limit levels, For the set of blocking actions, To assemble at the turnaround station, For resource allocation and collection, This indicates a combined enumeration of candidate actions. In the above formula, the speed limit level set... Blocking action set .

[0100] The above scheme generates scheduling strategy objects based on the updated disaster risk field, scheduling strategy levels based on risk probabilities, and scheduling strategy ranges based on system losses. A scheduling strategy template is then generated based on the scheduling strategy objects, scheduling strategy levels, and scheduling strategy ranges. Candidate segments are determined from the rail transit system based on risk probabilities, and candidate actions for these segments are determined from the scheduling strategy template. These candidate actions are then combined to obtain candidate scheduling strategies. Thus, each candidate scheduling strategy contains candidate actions and satisfies the scheduling strategy template, enabling the generation of a second strategy set based on these candidate scheduling strategies.

[0101] In some embodiments, step 104 includes: Step 1041: Determine multiple strategy combinations based on the first strategy set and the second strategy set.

[0102] In practice, the dynamic game model is determined at the current stage. The stage state, , in, For the dynamic game model at the current stage The stage state, This is the first twin state. This is the second twin state. For the current stage The comprehensive early warning index, For the current stage The set of safety constraints.

[0103] Phase State This includes the lockdown status, vehicle locations, available turnaround lines, emergency resource locations, and dispatch rules. Phase Status It is used to determine the set of optional actions, calculate the first loss function of the first twin, calculate the second loss function of the second twin, update the state of the next stage, determine whether the safety constraints are satisfied, and serve as the input to the objective function.

[0104] The first scheduling strategy is determined from the first strategy set. The first scheduling strategy is the first twin's strategy in the current stage. The output operation-side strategy. The first scheduling strategy is expressed as: , in, For the first twin at the current stage The first scheduling policy output, The first set of strategies (normal strategy set) for the first twin. This indicates that the system is operating normally. This indicates that the speed is limited. This indicates an adjustment to the tracking interval. Indicates a partial reversal. This indicates capacity compensation.

[0105] A second scheduling strategy is determined from the second strategy set. This second scheduling strategy is the strategy of the second twin in the current stage. The output emergency response strategy. The second scheduling strategy is expressed as: , in, For the second twin at the current stage The output second scheduling strategy, This is the second set of strategies (emergency strategy set) for the second twin. Indicates risk monitoring, Indicates interval blocking. This indicates that the guests have been cleared out and evacuated. This indicates the allocation of emergency resources. This indicates an emergency shutdown.

[0106] Multiple strategy combinations are determined based on the first scheduling strategy and the second scheduling strategy. ,in, For the first twin at the current stage The first scheduling policy output, For the second twin at the current stage The second scheduling strategy output.

[0107] Step 1042: Determine the first loss function of the first twin based on the first scheduling strategy in the first strategy set, and determine the second loss function of the second twin based on the second scheduling strategy in the second strategy set.

[0108] In practice, the process of determining the first loss function includes: Step 1042A: Obtain the upper limit of the reference delay time for rail transit and the expected delay time under the first scheduling strategy; determine the traffic delay loss based on the upper limit of the reference delay time and the expected delay time. ,in, For losses due to traffic delays, This represents the estimated delay time for rail transit under the first scheduling strategy. This is the upper limit of the reference delay time for rail transit.

[0109] Step 1042B: Obtain the reference passenger flow limit for rail transit and the excess passenger flow under the first scheduling strategy; determine the passenger backlog loss based on the reference passenger flow limit and the excess passenger flow. ,in, To avoid losses due to passenger backlog, For the excess passenger flow of rail transit under the first scheduling strategy, This represents the upper limit of passenger flow for rail transit. In the above formula, exceeding the passenger flow... This could be passenger flow exceeding the capacity of a station or section of the line.

[0110] Step 1042C: Obtain the basic energy consumption of the rail transit and the expected energy consumption under the first scheduling strategy; determine the energy consumption deviation loss based on the basic energy consumption and the expected energy consumption. ,in, For energy consumption deviation loss, This represents the projected energy consumption of rail transit under the first scheduling strategy. As the basic energy consumption of rail transit, This serves as a reference energy consumption limit for rail transit.

[0111] Step 1042D: Obtain the passenger's reference waiting time limit and the expected waiting time under the first scheduling strategy; determine the waiting time loss based on the reference waiting time limit and the expected waiting time. ,in, Due to the loss of waiting time, The expected waiting time under the first scheduling strategy. This serves as a reference waiting time limit for passengers.

[0112] The congestion loss, transfer delay loss, and service anomaly loss of passengers under the first scheduling strategy are obtained. The service quality loss is obtained by weighting the waiting time loss, congestion loss, transfer delay loss, and service anomaly loss. ,in, For service quality loss, The first weighting coefficient corresponding to the waiting time loss. To account for the passenger waiting time loss under the first scheduling strategy, The second weighting coefficient corresponds to the crowding level loss. To mitigate the loss of crowding levels on platforms or in carriages under the first scheduling strategy, The third weighting coefficient corresponding to the transfer delay loss. To compensate for losses due to failed transfers or transfer delays under the first scheduling strategy, The fourth weighting coefficient corresponding to service anomaly loss. This is to account for service disruptions experienced by passengers under the first scheduling strategy.

[0113] Step 1042E: Obtain the time deviation loss, interval loss, turnaround deviation loss, and train cancellation loss of rail transit under the first scheduling strategy. Then, perform weighted summation on these losses to obtain the operational stability loss. ,in, To ensure operational stability and avoid losses, This is the first weighting coefficient corresponding to the time deviation loss. This refers to the time deviation loss of train arrival or departure times in rail transit under the first scheduling strategy. The second weighting coefficient corresponds to the running interval loss. To account for the train interval loss under the first scheduling strategy, The third weighting coefficient corresponding to the return deviation loss. To account for the turnaround deviation loss of rail transit under the first scheduling strategy, The fourth weighting coefficient corresponding to the loss from train cancellations. This refers to the losses incurred due to train cancellations under the first scheduling strategy.

[0114] Step 1042F involves normalizing the train delay loss, passenger backlog loss, energy consumption deviation loss, service quality loss, and operational stability loss to avoid inconsistencies in the dimensions of each loss. The normalized train delay loss, passenger backlog loss, energy consumption deviation loss, service quality loss, and operational stability loss are then weighted to obtain the first loss function of the first twin. , in, The first loss function of the first twin. For the dynamic game model at the current stage The stage state, For the first twin at the current stage The first scheduling policy output, This is the first weighting coefficient corresponding to traffic delay losses. For normalized traffic delay losses, The second weighting coefficient corresponding to passenger backlog loss. For the normalized passenger backlog loss, The third weighting coefficient corresponding to the energy consumption deviation loss. This represents the normalized energy consumption deviation loss. The fourth weighting coefficient corresponding to the service quality loss. For the normalized service quality loss, The fifth weighting coefficient corresponding to the stable loss during operation. This represents the loss due to operational stability after normalization.

[0115] The process of determining the second loss function includes: Step 1042a: Determine the accident risk loss based on the risk probability of rail transit under the second scheduling strategy. ,in, For accident risk losses, For the set of affected risk areas, For line section Risk weights, For line section At the current stage The probability of risk.

[0116] Step 1042b: Obtain the passenger's exposure loss, evacuation time loss, congestion risk loss, and environmental intensity loss under the second scheduling strategy. Weight the exposure loss, evacuation time loss, congestion risk loss, and environmental intensity loss to obtain the passenger safety loss. ,in, For passenger safety losses, The first weighting coefficient corresponding to the exposure loss. To mitigate the loss of passenger exposure under the second scheduling strategy, The second weighting coefficient corresponds to the evacuation time loss. To account for the passenger evacuation time loss under the second scheduling strategy, The third weighting coefficient corresponding to the congestion risk loss. To mitigate passenger congestion risk losses under the second scheduling strategy, This is the fourth weighting coefficient corresponding to the environmental intensity loss. The environmental intensity loss for passengers under the second scheduling strategy.

[0117] Step 1042c: Obtain the normal service capacity of the rail transit system and the projected service capacity under the second scheduling strategy; determine the system stability loss based on the normal service capacity and the projected service capacity. ,in, For system stability loss, The projected service capacity of rail transit under the second dispatching strategy. This refers to the normal service capacity of rail transit. In the above formula, the projected service capacity is... It can be the remaining service capacity of rail transit under disaster conditions.

[0118] Step 1042d: Obtain the upper limit of the reference recovery time for rail transit and the estimated recovery time under the second scheduling strategy; determine the traffic recovery loss based on the upper limit of the reference recovery time and the estimated recovery time. ,in, For the losses caused by traffic restoration, This represents the estimated recovery time for rail transit under the second dispatching strategy. This serves as the upper limit for the reference recovery time of rail transit.

[0119] Step 1042e obtains the required resource quantity for rail transit and the available resource quantity under the second scheduling strategy, and determines the emergency resource loss based on the required and available resource quantities. ,in, To prevent the loss of emergency resources, This represents the amount of available resources for rail transit under the second scheduling strategy. The amount of resources required for rail transit.

[0120] Step 1042f involves normalizing the accident risk loss, passenger safety loss, system stability loss, traffic recovery loss, and emergency resource loss to avoid inconsistencies in the dimensions of each loss. The normalized accident risk loss, passenger safety loss, system stability loss, traffic recovery loss, and emergency resource loss are then weighted to obtain the second loss function of the second twin. , in, The second loss function of the second twin. For the dynamic game model at the current stage The stage state, For the second twin at the current stage The output second scheduling strategy, This is the first weighting coefficient corresponding to the accident risk loss. For the normalized accident risk loss, The second weighting coefficient corresponding to passenger safety losses. For normalized passenger safety losses, The third weighting coefficient corresponds to the system stability loss. The normalized system stability loss, The fourth weighting coefficient corresponding to the traffic recovery loss. For normalized traffic recovery losses, The fifth weighting coefficient corresponding to emergency resource losses. This refers to the normalized loss of emergency resources.

[0121] Step 1043: Determine a first optimal response strategy from the first strategy set based on the first loss function, and determine a second optimal response strategy from the second strategy set based on the second loss function.

[0122] In practical implementation, to ensure consistency between the game definition and the optimization of overall loss, this embodiment adopts a constrained cooperative game approach. The first twin and the second twin each determine their corresponding optimal response strategies, and under hard safety constraints, determine the equilibrium strategy combination that minimizes the overall loss and requires neither the first nor the second twin to deviate independently.

[0123] The first optimal response strategy is determined from the first strategy set based on the first loss function. , in, In order to implement a second scheduling policy The first optimal response strategy at that time The first scheduling policy value is chosen to minimize the first loss function. The first candidate strategy in the first strategy set. For the first set of strategies, This is the first loss function when considering the second scheduling strategy.

[0124] The second optimal response strategy is determined from the second strategy set based on the second loss function. , in, In order to implement a given first scheduling policy The second best response strategy at that time To determine the value of the second scheduling strategy that minimizes the second loss function, The second candidate policy in the second policy set. For the second set of strategies, This is the second loss function when considering the first scheduling strategy.

[0125] Step 1044: Using a dynamic game model, determine the target scheduling strategy for the rail transit from the multiple strategy combinations based on the first optimal response strategy and the second optimal response strategy.

[0126] In practice, a dynamic game model is used to determine whether an equilibrium strategy combination exists among multiple strategy combinations based on the first and second optimal response strategies. If an equilibrium strategy combination exists, the target scheduling strategy is determined from the equilibrium strategy combinations based on the objective function. If no equilibrium strategy combination exists, the target scheduling strategy that satisfies the preset deviation condition is determined from the multiple strategy combinations.

[0127] Figure 4 This is a schematic diagram of the solution space of the dynamic game model according to an embodiment of this disclosure. Figure 4As shown, scheduling strategies A, B, G, and H are candidate strategies. Strategies A and B are located in the double-loss region, while strategies G and H are located in the normal priority region. Strategies C, D, E, and F are Pareto front points, with strategy E being a Nash equilibrium point.

[0128] Figure 5 This is a flowchart of a rail transit control method based on a digital twin, according to an embodiment of this disclosure. Figure 5 As shown, the rail transit control method based on digital twins includes: During the normalized operation phase (dominated by the normalized twin), judgments are made on tiered early warnings; if the tiered early warning does not reach the threshold, the normalized operation phase is maintained; if the tiered early warning reaches the threshold, the disaster twin performs simulation and prediction (impact range trend / risk); emergency response alternatives (Solution 1, Solution 2, Solution 3, etc.) are generated; dynamic game decision-making (Nash equilibrium / optimal strategy) determines the target scheduling strategy; scheduling instructions are executed; real-time monitoring of weather / risk determines whether the emergency has subsided; if the emergency has not subsided, the disaster twin simulation and prediction are maintained; if the emergency subsides, a relocation transition plan is generated, operations are gradually restored, and the normalized twin takes over.

[0129] The above scheme determines multiple strategy combinations based on a first strategy set and a second strategy set. A first loss function for the first twin is determined based on the first scheduling strategy in the first strategy set, and a second loss function for the second twin is determined based on the second scheduling strategy in the second strategy set. A first optimal response strategy is determined from the first strategy set based on the first loss function, and a second optimal response strategy is determined from the second strategy set based on the second loss function. Using a dynamic game model, a target scheduling strategy for rail transit is determined from multiple strategy combinations based on the first and second optimal response strategies. In this way, the target scheduling strategy can simultaneously consider both routine operational efficiency and extreme weather safety risks, thereby comprehensively improving the stability and safety of rail transit operation.

[0130] In some embodiments, step 1044 includes: Step 10441: Determine the objective function of the dynamic game model based on the first loss function and the second loss function.

[0131] In practice, the trade-off coefficients are determined based on the comprehensive early warning index and the risk probability. , in, For the current stage The trade-off coefficient, As the benchmark trade-off coefficient, The first adjustment factor is... For the current stage The comprehensive early warning index, This is the second adjustment factor. This represents the average risk probability of the line segment. This is the third adjustment factor. For passenger exposure.

[0132] In the above formula, the higher the warning level, the greater the average risk level, and the higher the passenger exposure, the higher the tradeoff coefficient. The larger the value, the more the dynamic game model tends to favor the safety and emergency response goals.

[0133] The objective function of the dynamic game model is determined based on the first loss function and the second loss function. , in, Let be the objective function of the dynamic game model. The goal is to find the minimum overall loss among the combinations of the first and second scheduling strategies. For the first twin at the current stage The first scheduling policy output, For the second twin at the current stage The output second scheduling strategy, The first loss function of the first twin. For the current stage The trade-off coefficient, The second loss function of the second twin. As a discount factor, For the next stage In the next state The objective function is as follows.

[0134] The objective function is used to evaluate the first scheduling strategy. Combination with the second scheduling strategy The cumulative overall loss incurred in the current stage and within the future prediction window is the objective function of the dynamic game model for strategy comparison, recursive solution, and final selection of the scheduling strategy. The dynamic game model minimizes this objective function to select the equilibrium strategy combination with the minimum overall loss while satisfying hard safety constraints.

[0135] Step 10442: Based on the first optimal response strategy and the second optimal response strategy, determine whether there is a balanced strategy combination among the multiple strategy combinations.

[0136] In practice, at the current stage When strategy combination If neither twin can reduce the loss function by unilaterally changing the output scheduling strategy, then the strategy combination is determined to be the Nash equilibrium strategy combination for the current stage. ,in, For strategy combination, This is a Nash equilibrium strategy combination.

[0137] Specifically, when strategy combination Simultaneously satisfying the condition that the first scheduling strategy belongs to the first optimal response strategy and the second scheduling strategy belongs to the second optimal response strategy, that is... and Then determine the strategy combination. This represents a combination of Nash equilibrium strategies. The above conditions indicate that both the first and second scheduling strategies are the optimal response strategies under their respective conditions.

[0138] Step 10443: In response to the existence of a balanced strategy combination among the multiple strategy combinations, a target scheduling strategy is determined from the balanced strategy combination based on the objective function.

[0139] In practice, the current stage When multiple equilibrium strategy combinations exist, the objective function is determined from these combinations. The target scheduling strategy that minimizes the set of safety constraints.

[0140] Step 10444: In response to the absence of a balanced strategy combination among the multiple strategy combinations, a target scheduling strategy that satisfies the preset deviation condition is determined from the multiple strategy combinations.

[0141] In practice, when no balanced strategy combination exists among multiple strategy combinations, a target scheduling strategy that meets a preset deviation condition is determined from the multiple strategy combinations. The preset deviation condition means that the deviation between the strategy combination and the balanced strategy combination is less than a preset deviation threshold.

[0142] At the current stage First, fix the second scheduling strategy and solve for the first optimal response strategy of the first twin; then fix the first scheduling strategy and solve for the second optimal response strategy of the second twin; alternately update and verify the set of safety constraints until the target scheduling strategy is stable, the change in the objective function is less than the preset change threshold, or the maximum number of iterations is reached. If convergence is not achieved after reaching the maximum number of iterations, the system reverts to the safety-first strategy, that is, selects the target scheduling strategy with the smallest second loss function that satisfies the set of safety constraints, and prompts the scheduler for manual confirmation.

[0143] The above scheme determines the objective function of the dynamic game model based on the first and second loss functions. Based on the first and second optimal response strategies, it is determined whether an equilibrium strategy combination exists among multiple strategy combinations. If an equilibrium strategy combination exists, a target scheduling strategy is determined from this combination based on the objective function. This method of determining the target scheduling strategy from the equilibrium strategy combination simultaneously considers both normal operational efficiency and extreme weather safety risks. If no equilibrium strategy combination exists, a target scheduling strategy that satisfies a preset deviation condition is determined from the multiple strategy combinations. This method of ensuring the target scheduling strategy satisfies the preset deviation condition maximizes the balance between normal operational efficiency and extreme weather safety risks.

[0144] In some embodiments, step 10444 includes: Step 10444A: Determine a candidate strategy combination from the plurality of strategy combinations, wherein the candidate strategy combination includes: a first candidate strategy and a second candidate strategy.

[0145] In practice, a candidate strategy combination is determined from multiple strategy combinations, including a first candidate strategy and a second candidate strategy. This facilitates the determination of the degree of deviation between the first candidate strategy, the second candidate strategy, and the equilibrium strategy combination.

[0146] Step 10444B: Determine the first deviation between the first candidate strategy and the first optimal response strategy, determine the second deviation between the second candidate strategy and the second optimal response strategy, and determine the target deviation based on the first deviation and the second deviation.

[0147] In practice, the target deviation is obtained by summing the first deviation and the second deviation, or by averaging the first deviation and the second deviation. The target deviation reflects the overall degree of deviation between the candidate strategy combination and the equilibrium strategy combination.

[0148] Step 10444C: Obtain the basic deviation coefficient, and determine the preset deviation threshold based on the comprehensive early warning index, risk probability, and the basic deviation coefficient.

[0149] In practice, the preset deviation threshold is determined based on the comprehensive early warning index, risk probability, and basic deviation coefficient. , in, To set a preset deviation threshold, The basic deviation coefficient, The first adjustment coefficient, For the current stage The comprehensive early warning index, This is the second adjustment coefficient. This represents the average risk probability of the line segment.

[0150] Step 10444D: In response to the target deviation being less than the preset deviation threshold, it is determined that the candidate strategy combination meets the preset deviation condition, and the candidate strategy combination is then used as the target strategy combination.

[0151] In practice, when the target deviation is less than the preset deviation threshold If the candidate strategy combination is determined to meet the preset deviation condition, then the candidate strategy combination will be used as the target strategy combination.

[0152] The above scheme determines candidate strategy combinations from multiple strategy combinations, including a first candidate strategy and a second candidate strategy. A first deviation between the first candidate strategy and the first optimal response strategy is determined, and a second deviation between the second candidate strategy and the second optimal response strategy is determined. A target deviation is determined based on the first and second deviations. A basic deviation coefficient is obtained, and a preset deviation threshold is determined based on the comprehensive early warning index, risk probability, and the basic deviation coefficient. If the target deviation is less than the preset deviation threshold, the candidate strategy combination is deemed to meet the preset deviation condition and is then adopted as the target strategy combination. In this way, by determining the target strategy combination with the smallest deviation from the equilibrium strategy combination from multiple strategy combinations, it is possible to maximize both normal operational efficiency and extreme weather safety risks.

[0153] Through the above embodiments, operational and environmental data of the rail transit are acquired. Using a first twin, a first set of strategies that meets the operational efficiency requirements of the rail transit is determined based on the operational and environmental data; wherein, the first set of strategies is the set of scheduling strategies for the rail transit under normal conditions. Using a second twin, a second set of strategies that meets the safety risks of the rail transit is determined based on the operational and environmental data; wherein, the second set of strategies is the set of scheduling strategies for the rail transit under disaster conditions. Thus, by using the first and second twins, respectively outputting a first set of strategies adapted to normal operational efficiency and a second set of strategies adapted to extreme weather safety risks, the problem that a single twin architecture cannot simultaneously address both normal operation and safety risks can be solved. Using a dynamic game theory model, a target scheduling strategy for the rail transit is determined based on the first and second strategy sets, and the rail transit is controlled based on the target scheduling strategy. Thus, by using a dynamic game theory model to collaboratively deduce the first and second strategy sets, a target scheduling strategy that balances efficiency and safety can be accurately output, thereby comprehensively improving the stability and safety of rail transit operation.

[0154] During the dissipation of extreme weather, the system does not immediately return to normal but enters a mixed migration phase. The migration determination index is defined as: , in, For the current stage The migration determination index, This is the first weighting coefficient corresponding to the comprehensive early warning index. As a comprehensive early warning index, This is the second weighting coefficient corresponding to the probability of accident risk. This represents the normalized probability of accident risk. This is the third weighting coefficient corresponding to the system resilience loss. This represents the normalized system resilience loss. This is the fourth weighting coefficient corresponding to the anomaly rate of critical equipment re-inspection. For the abnormality rate of key equipment re-inspection. When the relocation judgment index continuous When a time window is lower than the backtracking threshold, the system gradually transfers the leadership of the second twin to the first twin.

[0155] The transfer of control is divided into four phases: second twin dominance, hybrid dominance, first twin enhancement, and first twin dominance again. In the second twin dominance phase, safety loss carries a higher weight, and the second twin maintains high-frequency simulations. In the hybrid dominance phase, the system allows the first twin to propose strategies such as increasing speed limits, restoring some capacity, and optimizing the operational schedule. In the first twin enhancement phase, the second twin shifts from full-speed simulation to focused monitoring. In the first twin dominance phase, the system restores first twin dominance, and the second twin enters a low-power warning scanning state. If the migration judgment index rises again, the accident risk probability rebounds, or equipment re-inspection fails during the migration process, the migration is suspended or the system reverts to the second twin dominance phase.

[0156] Strategy self-learning is achieved by comparing predicted risk with actual feedback risk. The updated tradeoff coefficients are: , in, For the next stage The trade-off coefficient, For the current stage The trade-off coefficient, For learning rate, The observed risks obtained from actual implementation feedback, To predict risk for the model, To weigh the lower limit of the coefficient, To balance the upper limit of the coefficient, To limit the updated tradeoff coefficient to Internally, avoid unbounded growth or negative values ​​in weights. When actual risk is higher than predicted risk, increase the trade-off coefficient in subsequent decisions; when actual risk is lower than predicted risk and operational losses are significantly higher, decrease the overly conservative trade-off coefficient.

[0157] Figure 6 This is a schematic diagram comparing the implementation effects of the embodiments of this disclosure. Figure 6 As shown, the normal operational efficiency retention rate of the single twin solution is 45%, while the normal operational efficiency retention rate of the dual-twin combined game decision-making solution in this embodiment is 92%. Compared to the single twin solution, this embodiment has a higher normal operational efficiency retention rate. The emergency response delay of the single twin solution is 35 minutes, while the emergency response delay of the dual-twin combined game decision-making solution in this embodiment is 8 minutes, showing a shorter emergency response delay. The decision-making scheme generation time of the single twin solution is 180 seconds, while the decision-making scheme generation time of the dual-twin combined game decision-making solution in this embodiment is 25 seconds, also showing a shorter generation time. The mode switching smoothness of the single twin solution is 40 points, while the mode switching smoothness of the dual-twin combined game decision-making solution in this embodiment is 95 points, demonstrating higher mode switching smoothness compared to the single twin solution. The overall security accuracy of the single twin scheme is 62%, while the overall security accuracy of the dual twin combined game decision-making scheme in this embodiment is 96%. Compared with the single twin scheme, the embodiment of this disclosure has higher overall security accuracy.

[0158] It should be noted that the method of this disclosure embodiment can be executed by a single device, such as a computer or server. The method of this embodiment can also be applied to a distributed scenario, where multiple devices cooperate to complete the task. In such a distributed scenario, one of these devices may execute only one or more steps of the method of this disclosure embodiment, and the multiple devices will interact with each other to complete the method described.

[0159] It should be noted that the above description describes some embodiments of this disclosure. Other embodiments are within the scope of the appended claims of this disclosure. In some cases, the actions or steps described in the claims of this disclosure may be performed in a different order than that shown in the above embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0160] Based on the same inventive concept, corresponding to any of the above-described embodiments, this disclosure also provides a control device for rail transit.

[0161] refer to Figure 7 The control device for the rail transit includes: The acquisition module 301 is configured to acquire the operation data and environmental data of the rail transit system. The first strategy set determination module 302 is configured to use a first twin to determine a first strategy set that satisfies the operational efficiency of rail transit based on the operational data and the environmental data; wherein, the first strategy set is the scheduling strategy set of the rail transit under normal conditions. The second strategy set determination module 303 is configured to use a second twin to determine a second strategy set that meets the rail transit safety risk based on the operational data and the environmental data; wherein, the second strategy set is the scheduling strategy set of the rail transit under disaster conditions; The target scheduling strategy determination module 304 is configured to use a dynamic game model to determine the target scheduling strategy of the rail transit based on the first strategy set and the second strategy set, and to control the rail transit based on the target scheduling strategy.

[0162] In some embodiments, the second strategy set determination module 303 includes: The meteorological hazard index determination unit is configured to acquire real-time meteorological values ​​and safety reference values ​​of various meteorological indicators, and determine the meteorological hazard index based on the real-time meteorological values ​​and the safety reference values. The line hazard index determination unit is configured to acquire risk coefficients of multiple lines and perform weighted processing on the risk coefficients of the multiple lines to obtain the line hazard index. The indicator anomaly index determination unit is configured to acquire real-time monitoring values ​​and historical normal averages of multiple monitoring indicators, and determine the indicator anomaly index based on the real-time monitoring values ​​and the historical normal averages. The historical disaster index determination unit is configured to acquire frequency parameters, impact parameters, average recovery time, and early warning error correction values ​​of historical disasters, and to perform weighted processing on the frequency parameters, impact parameters, average recovery time, and early warning error correction values ​​to obtain the historical disaster index; The comprehensive early warning index determination unit is configured to perform weighted processing on the meteorological hazard index, the line hazard index, the indicator anomaly index, and the historical disaster index to obtain the comprehensive early warning index; The operating frequency determination unit is configured to determine the warning level based on the comprehensive warning index and to determine the operating frequency based on the warning level. The second scheduling strategy determination unit is configured to invoke the second twin based on the operating frequency, and use the second twin to determine the second strategy set based on the operating data and the environmental data.

[0163] In some embodiments, the second scheduling policy determination module 303 includes: The disaster risk field determination unit is configured to acquire current weather data and a rail transit line topology map, and determine and update the disaster risk field based on the current weather data and the line topology map; The risk probability determination unit is configured to acquire weather intensity indicators, disaster environment indicators, train operating speed, section hazard coefficient, and passenger parameters, and determine the risk probability of rail transit based on the weather intensity indicators, the disaster environment indicators, the train operating speed, the section hazard coefficient, and the passenger parameters. The system loss determination unit is configured to acquire a first functional parameter of the rail transit under disaster conditions and a second functional parameter under normal conditions, and determine the system loss of the rail transit based on the first functional parameter and the second functional parameter. The candidate scheduling strategy generation unit is configured to generate a scheduling strategy template based on the updated disaster risk field, the risk probability, and the system loss, and to generate candidate scheduling strategies based on the scheduling strategy template. The second scheduling strategy determination unit is configured to retrieve a set of security constraints and determine a second strategy set based on the candidate scheduling strategies and the set of security constraints.

[0164] In some embodiments, the candidate scheduling policy generation unit includes: The scheduling strategy parameter determination subunit is configured to generate a scheduling strategy object based on the updated disaster risk field, generate a scheduling strategy level based on the risk probability, and generate a scheduling strategy range based on the system loss. The scheduling strategy template generation subunit is configured to generate a scheduling strategy template based on the scheduling strategy object, the scheduling strategy level, and the scheduling strategy range. The candidate scheduling strategy generation subunit is configured to determine candidate segments from the rail transit based on the risk probability, determine candidate actions for the candidate segments from the scheduling strategy template, and combine the candidate actions to obtain a candidate scheduling strategy.

[0165] In some embodiments, the target scheduling strategy determination module 304 includes: The strategy combination determination unit is configured to determine multiple strategy combinations based on the first strategy set and the second strategy set; The loss function determination unit is configured to determine a first loss function of the first twin based on a first scheduling strategy in the first strategy set, and to determine a second loss function of the second twin based on a second scheduling strategy in the second strategy set. The optimal response strategy determination unit is configured to determine a first optimal response strategy from the first strategy set based on the first loss function, and to determine a second optimal response strategy from the second strategy set based on the second loss function; The target scheduling strategy determination unit is configured to use a dynamic game model to determine the target scheduling strategy of the rail transit from the multiple strategy combinations based on the first optimal response strategy and the second optimal response strategy.

[0166] In some embodiments, the target scheduling policy determination unit includes: The objective function determination sub-unit is configured to determine the objective function of the dynamic game model based on the first loss function and the second loss function; The balanced strategy combination judgment subunit is configured to determine whether there is a balanced strategy combination among the multiple strategy combinations based on the first best response strategy and the second best response strategy. The first target scheduling strategy determination subunit is configured to determine a target scheduling strategy from the balanced strategy combination based on the objective function in response to the existence of a balanced strategy combination among the plurality of strategy combinations. The second target scheduling strategy determination subunit is configured to determine a target scheduling strategy that satisfies a preset deviation condition from the multiple strategy combinations if there is no balanced strategy combination among the multiple strategy combinations.

[0167] In some embodiments, the second target scheduling policy determining subunit is specifically configured as follows: From the plurality of strategy combinations, a candidate strategy combination is determined, wherein the candidate strategy combination includes: a first candidate strategy and a second candidate strategy; Determine a first deviation between the first candidate strategy and the first optimal response strategy, determine a second deviation between the second candidate strategy and the second optimal response strategy, and determine a target deviation based on the first deviation and the second deviation; Obtain the basic deviation coefficient, and determine the preset deviation threshold based on the comprehensive early warning index, risk probability, and the basic deviation coefficient; If the target deviation is less than the preset deviation threshold, it is determined that the candidate strategy combination meets the preset deviation condition, and the candidate strategy combination is then used as the target strategy combination.

[0168] For ease of description, the above apparatus is described in terms of its functions, divided into various modules. Of course, in implementing this disclosure, the functions of each module can be implemented in one or more software and / or hardware.

[0169] The apparatus of the above embodiments is used to implement the corresponding rail transit control method in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0170] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this disclosure also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the rail transit control method described in any of the above embodiments.

[0171] Figure 8 This embodiment illustrates a more specific hardware structure of an electronic device, which may include a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, memory 1020, input / output interface 1030, and communication interface 1040 are interconnected internally via the bus 1050.

[0172] The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0173] The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program code is stored in the memory 1020 and is called and executed by the processor 1010.

[0174] The input / output interface 1030 is used to connect input / output modules to realize information input and output. Input / output modules can be configured as components within the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touchscreens, microphones, various sensors, etc., while output devices may include displays, speakers, vibrators, indicator lights, etc.

[0175] The communication interface 1040 is used to connect the communication module (not shown in the figure) to enable communication between this device and other devices. The communication module can communicate via wired means (such as USB (Universal Serial Bus), network cable, etc.) or wireless means (such as mobile network, WIFI (Wireless Fidelity), Bluetooth, etc.).

[0176] Bus 1050 includes a pathway for transmitting information between various components of the device, such as processor 1010, memory 1020, input / output interface 1030, and communication interface 1040.

[0177] It should be noted that although the above-described device only shows the processor 1010, memory 1020, input / output interface 1030, communication interface 1040, and bus 1050, in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the embodiments of this specification, and not necessarily all the components shown in the figures.

[0178] The electronic devices described above are used to implement the corresponding rail transit control methods in any of the foregoing embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0179] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this disclosure also provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the rail transit control method as described in any of the above embodiments.

[0180] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0181] The computer instructions stored in the storage medium of the above embodiments are used to cause the computer to execute the rail transit control method as described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0182] Based on the same inventive concept, corresponding to any of the above embodiments, this application also provides a computer program product, including computer program instructions. When the computer program instructions are run on a computer, the computer causes the computer to execute the rail transit control method as described in any of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0183] It is understood that before using the technical solutions of the various embodiments in this disclosure, users will be informed of the type, scope of use, and usage scenarios of the personal information involved in an appropriate manner, and user authorization will be obtained.

[0184] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose, based on the prompt message, whether to provide personal information to the software or hardware such as electronic devices, applications, servers, or storage media performing the operations of this disclosed technical solution.

[0185] As an optional but not limited implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0186] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this disclosure.

[0187] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of this disclosure is limited to these examples; within the framework of this disclosure, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of the embodiments of this disclosure as described above, which are not provided in detail for the sake of brevity.

[0188] Additionally, to simplify the description and discussion, and to avoid obscuring the embodiments of this disclosure, the provided drawings may or may not show well-known power / ground connections to integrated circuit (IC) chips and other components. Furthermore, the apparatus may be shown in block diagram form to avoid obscuring the embodiments of this disclosure, and this also takes into account the fact that the details of implementation of these block diagram apparatuses are highly dependent on the platform on which the embodiments of this disclosure will be implemented (i.e., these details should be fully understood by those skilled in the art). While specific details (e.g., circuits) have been set forth to describe exemplary embodiments of this disclosure, it will be apparent to those skilled in the art that the embodiments of this disclosure can be implemented without these specific details or with variations thereof. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0189] Although this disclosure has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0190] This disclosure is intended to cover all such substitutions, modifications, and variations that fall within the broad scope of this disclosure. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this disclosure should be included within the protection scope of this disclosure.

Claims

1. A control method for rail transit, characterized in that, The method includes: Acquire operational and environmental data of rail transit; Using a first twin, a first set of strategies that meet the operational efficiency of rail transit is determined based on the operational data and the environmental data; wherein, the first set of strategies is the set of scheduling strategies when the rail transit is in a normal state; Using a second twin, a second set of strategies that meet the safety risks of rail transit is determined based on the operational data and the environmental data; wherein, the second set of strategies is a set of scheduling strategies for rail transit under disaster conditions; Using a dynamic game model, a target scheduling strategy for the rail transit is determined based on the first strategy set and the second strategy set, and the rail transit is controlled based on the target scheduling strategy. The method of using a second twin to determine a second set of strategies that meet the requirements of rail transit safety risks based on the operational data and the environmental data includes: Obtain current weather data and a rail transit route topology map, and determine and update disaster risk fields based on the current weather data and the route topology map; The system acquires weather intensity indicators, disaster environment indicators, train operating speed, section hazard coefficient, and passenger parameters, and determines the risk probability of rail transit based on these parameters. Obtain the first functional parameters of rail transit under disaster conditions and the second functional parameters under normal conditions, and determine the system loss of rail transit based on the first functional parameters and the second functional parameters; A scheduling strategy template is generated based on the updated disaster risk field, the risk probability, and the system loss; and a candidate scheduling strategy is generated based on the scheduling strategy template. Retrieve the set of security constraints, and determine a second set of strategies based on the candidate scheduling strategies and the set of security constraints; The set of safety constraints includes: block safety constraints, interlocking safety constraints, platform capacity constraints, passenger evacuation capacity constraints, emergency resource arrival time constraints, and maximum permissible risk constraints. The process of generating a scheduling strategy template based on the updated disaster risk field, the risk probability, and the system loss, and generating candidate scheduling strategies based on the scheduling strategy template, includes: A scheduling strategy object is generated based on the updated disaster risk field, a scheduling strategy level is generated based on the risk probability, and a scheduling strategy range is generated based on the system loss. A scheduling policy template is generated based on the scheduling policy object, the scheduling policy level, and the scheduling policy range; Candidate segments are determined from the rail transit system based on the risk probability, candidate actions for the candidate segments are determined from the scheduling strategy template, and the candidate actions are combined to obtain a candidate scheduling strategy.

2. The method according to claim 1, characterized in that, The method of using a second twin to determine a second set of strategies that meet the requirements of rail transit safety risks based on the operational data and the environmental data includes: The system acquires real-time meteorological values ​​and safety reference values ​​for various meteorological indicators, and determines a meteorological hazard index based on the real-time meteorological values ​​and the safety reference values. Obtain the risk coefficients of multiple routes, and then perform a weighted average of the risk coefficients of the multiple routes to obtain the route hazard index; Obtain real-time monitoring values ​​and historical normal averages of various monitoring indicators, and determine the indicator abnormality index based on the real-time monitoring values ​​and the historical normal averages; The frequency parameters, impact parameters, average recovery time, and early warning error correction values ​​of historical disasters are obtained. The frequency parameters, impact parameters, average recovery time, and early warning error correction values ​​are weighted to obtain the historical disaster index. The meteorological hazard index, the line hazard index, the indicator anomaly index, and the historical disaster index are weighted to obtain a comprehensive early warning index. The warning level is determined based on the comprehensive warning index, and the operating frequency is determined based on the warning level. The second twin is invoked according to the operating frequency, and the second twin is used to determine the second strategy set based on the operating data and the environmental data.

3. The method according to claim 1, characterized in that, The step of determining the target scheduling strategy for the rail transit based on the first strategy set and the second strategy set using a dynamic game model includes: Multiple strategy combinations are determined based on the first strategy set and the second strategy set; The first loss function of the first twin is determined according to the first scheduling strategy in the first strategy set, and the second loss function of the second twin is determined according to the second scheduling strategy in the second strategy set; A first optimal response strategy is determined from the first strategy set based on the first loss function, and a second optimal response strategy is determined from the second strategy set based on the second loss function; Using a dynamic game model, the target scheduling strategy for rail transit is determined from the multiple strategy combinations based on the first optimal response strategy and the second optimal response strategy.

4. The method according to claim 3, characterized in that, The step of using a dynamic game model to determine the target scheduling strategy for rail transit from the multiple strategy combinations based on the first optimal response strategy and the second optimal response strategy includes: The objective function of the dynamic game model is determined based on the first loss function and the second loss function. Based on the first optimal response strategy and the second optimal response strategy, determine whether there is a balanced strategy combination among the multiple strategy combinations; If a balanced strategy combination exists among the plurality of strategy combinations, then a target scheduling strategy is determined from the balanced strategy combination based on the objective function. If no balanced strategy combination exists among the multiple strategy combinations, then a target scheduling strategy that satisfies the preset deviation condition is determined from the multiple strategy combinations.

5. The method according to claim 4, characterized in that, The step of determining the target scheduling strategy that satisfies the preset deviation condition from the multiple strategy combinations includes: From the plurality of strategy combinations, a candidate strategy combination is determined, wherein the candidate strategy combination includes: a first candidate strategy and a second candidate strategy; Determine a first deviation between the first candidate strategy and the first optimal response strategy, determine a second deviation between the second candidate strategy and the second optimal response strategy, and determine a target deviation based on the first deviation and the second deviation; Obtain the basic deviation coefficient, and determine the preset deviation threshold based on the comprehensive early warning index, risk probability, and the basic deviation coefficient; If the target deviation is less than the preset deviation threshold, it is determined that the candidate strategy combination meets the preset deviation condition, and the candidate strategy combination is then used as the target strategy combination.

6. A control device for rail transit, characterized in that, include: The acquisition module is configured to acquire operational and environmental data of the rail transit system. The first strategy set determination module is configured to use a first twin to determine a first strategy set that satisfies the operational efficiency of rail transit based on the operational data and the environmental data; wherein, the first strategy set is the scheduling strategy set of the rail transit under normal conditions; The second strategy set determination module is configured to use a second twin to determine a second strategy set that meets the rail transit safety risk based on the operational data and the environmental data; wherein, the second strategy set is the scheduling strategy set of the rail transit under disaster conditions; The target scheduling strategy determination module is configured to use a dynamic game model to determine the target scheduling strategy of the rail transit based on the first strategy set and the second strategy set, and to control the rail transit based on the target scheduling strategy. The second strategy set determination module includes: The disaster risk field determination unit is configured to acquire current weather data and a rail transit line topology map, and determine and update the disaster risk field based on the current weather data and the line topology map; The risk probability determination unit is configured to acquire weather intensity indicators, disaster environment indicators, train operating speed, section hazard coefficient, and passenger parameters, and determine the risk probability of rail transit based on the weather intensity indicators, the disaster environment indicators, the train operating speed, the section hazard coefficient, and the passenger parameters. The system loss determination unit is configured to acquire a first functional parameter of the rail transit under disaster conditions and a second functional parameter under normal conditions, and determine the system loss of the rail transit based on the first functional parameter and the second functional parameter. The candidate scheduling strategy generation unit is configured to generate a scheduling strategy template based on the updated disaster risk field, the risk probability, and the system loss, and to generate candidate scheduling strategies based on the scheduling strategy template. The second scheduling strategy determination unit is configured to retrieve a set of security constraints and determine a second strategy set based on the candidate scheduling strategies and the set of security constraints. The set of safety constraints includes: block safety constraints, interlocking safety constraints, platform capacity constraints, passenger evacuation capacity constraints, emergency resource arrival time constraints, and maximum permissible risk constraints. The candidate scheduling policy generation unit includes: The scheduling strategy parameter determination subunit is configured to generate a scheduling strategy object based on the updated disaster risk field, generate a scheduling strategy level based on the risk probability, and generate a scheduling strategy range based on the system loss. The scheduling strategy template generation subunit is configured to generate a scheduling strategy template based on the scheduling strategy object, the scheduling strategy level, and the scheduling strategy range. The candidate scheduling strategy generation subunit is configured to determine candidate segments from the rail transit based on the risk probability, determine candidate actions for the candidate segments from the scheduling strategy template, and combine the candidate actions to obtain a candidate scheduling strategy.

7. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor, when executing the program, implements the method as claimed in any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores computer instructions for causing a computer to perform the method described in any one of claims 1 to 5.

Citation Information

Patent Citations

  • Dynamic road network collaborative expansion decision system based on federated learning-digital twinning

    CN120877530A

  • Tourism resource scheduling method and device based on multi-modal data, equipment and storage medium

    CN121212742A

  • Intelligent water conservancy multi-mode perception and decision optimization system

    CN121836438A