A method, equipment, medium, and product for adjusting the operation of urban rail distributed trains.

Abstract

This application discloses a method, equipment, medium, and product for adjusting the operation of distributed trains in urban rail transit, relating to the field of train operation organization and control technology. The method includes: acquiring information data of the distributed train system in urban rail transit; constructing a train cooperative control model for sections with insufficient power supply based on the information data; solving the train cooperative control model using a hybrid barrier deep Q-network algorithm to determine the train cooperative control strategy for sections with insufficient power supply, thereby adjusting train operation; constructing a train timetable adjustment model for sections with sufficient power supply based on the train operation adjustment results for sections with insufficient power supply; solving the train timetable adjustment model using a preset solver, and combining it with the train cooperative control strategy for sections with insufficient power supply to obtain the adjusted train timetable. This application can realize train cooperative control for sections with insufficient power supply and adjustment of train arrival and departure times for sections with sufficient power supply, ensuring the feasibility and efficiency of the adjustment results.

Description

Technical Field

[0001] This application relates to the field of train operation organization and control technology, and in particular to a method, equipment, medium and product for adjusting distributed train operation in urban rail transit. Background Technology

[0002] In recent years, urban rail transit (hereinafter referred to as "urban rail") has developed rapidly due to its safety, energy saving, and high efficiency. The traction power supply system plays a crucial role as the "power heart," aiming to provide sufficient traction power for urban rail trains. However, due to factors such as equipment aging and foreign object intrusion, traction power supply failures frequently occur during daily operation. This leads to the malfunction of the DC feeder circuit breaker on one side of the traction substation, requiring temporary single-side power supply to maintain power for the entire line. During this period, due to traction power limitations, trains cannot apply traction conditions as planned. Consequently, train density and operating speed decrease, resulting in reduced line capacity and a significant impact on train operation and passenger services. Real-time train operation adjustments are necessary to mitigate the impact of these failures.

[0003] To address train operation adjustments in scenarios with insufficient unidirectional power supply, dispatchers currently primarily use train timetable adjustments to regulate train arrival and departure times. This method relies on the conservative assumption that all trains always apply maximum traction power. It significantly increases the headway in power-deficient sections according to dispatching procedures, thereby reducing train density in those sections and ensuring that even if all trains apply maximum traction power simultaneously, the total traction power will not exceed the rated range, thus preventing overload. However, current train operation adjustment methods for unidirectional power supply shortage scenarios have the following drawbacks: In actual operation, trains do not always apply maximum traction power; their actual power changes dynamically with the switching of driving conditions. This deviation makes it difficult for the current train timetable adjustment method to make full use of the limited power supply capacity, resulting in severely limited train operation efficiency.

[0004] Currently, train timetable methods can uniformly adjust the arrival and departure times of trains across the entire line using an overall model. However, after considering the dynamic power of trains, it is necessary to handle the adjustment issues of different sections using two methods: train cooperative control and train timetable adjustment, based on the power supply capacity of different sections. The overall model is no longer applicable, and a distributed train operation adjustment method is urgently needed.

[0005] The current manual adjustment method is difficult to take into account information from multiple disciplines such as train operation and power supply to achieve cross-disciplinary scheduling decisions, and it is difficult to guarantee the feasibility and efficiency of the adjustment results. Summary of the Invention

[0006] The purpose of this application is to provide a method, equipment, medium, and product for adjusting the operation of distributed trains in urban rail transit, which can realize the coordinated control of trains in sections with insufficient power supply and the adjustment of train arrival and departure times in sections with sufficient power supply, so as to ensure the feasibility and efficiency of the adjustment results.

[0007] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a method for adjusting the operation of urban rail distributed trains, which is applied to a scenario of insufficient unidirectional power supply; wherein, insufficient power supply means that the traction power provided by the traction power supply system is less than the operating power required by the urban rail train. The method for adjusting the operation of distributed trains on urban rail transit includes: Acquire information data of the urban rail distributed train system; the information data includes planned operation schedule information, traction power supply fault information, and basic line and operation data of urban rail trains; A train cooperative control model for the power shortage section is constructed based on the information data; the train cooperative control model includes a first objective function and a first constraint condition; The hybrid barrier deep Q-network algorithm is used to solve the train cooperative control model to determine the train cooperative control strategy for the power shortage section, so as to adjust the train operation and obtain the train operation adjustment result for the power shortage section. Based on the train operation adjustment results for sections with insufficient power supply, a train timetable adjustment model for sections with sufficient power supply is constructed. The train timetable adjustment model includes a second objective function and a second constraint condition. Sufficient power supply means that the traction power provided by the traction power supply system is not less than the operating power required by the urban rail train. The train timetable adjustment model is solved using a preset solver, and combined with the train cooperative control strategy for sections with insufficient power supply, to obtain the adjusted train timetable; the train timetable includes: train arrival and departure times and operating routes.

[0008] Secondly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the urban rail distributed train operation adjustment method described above.

[0009] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned urban rail distributed train operation adjustment method.

[0010] Fourthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the aforementioned urban rail distributed train operation adjustment method.

[0011] According to the specific embodiments provided in this application, the following technical effects are disclosed: This application provides a method, device, medium, and product for adjusting the operation of distributed trains in urban rail transit. It involves acquiring information data from the distributed train system; constructing a train cooperative control model for power-deficient sections based on the information data; solving the train cooperative control model using a hybrid barrier deep Q-network algorithm to determine the train cooperative control strategy for power-deficient sections, thereby adjusting train operation; constructing a train timetable adjustment model for power-sufficient sections based on the adjustment results; solving the train timetable adjustment model using a pre-set solver; and combining the train cooperative control strategy for power-deficient sections to obtain the adjusted train timetable. This application can implement train cooperative control for power-deficient sections using a hybrid barrier deep Q-network algorithm based on the train cooperative control model, and adjust train arrival and departure times for power-sufficient sections using a pre-set solver based on the train timetable adjustment model. Furthermore, the integration of the train cooperative control strategy for power-deficient sections during the train timetable adjustment process ensures the feasibility and efficiency of the adjustment results. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 A flowchart for adjusting the operation of distributed trains in urban rail transit; Figure 2 A flowchart illustrating the method for adjusting the operation of distributed trains in urban rail transit systems in scenarios with insufficient unidirectional power supply. Figure 3 This is a topology diagram of the urban rail transit line; Figure 4 A schematic diagram of the structure of the distributed train operation adjustment method; Figure 5 A schematic diagram of a safe reinforcement learning algorithm; Figure 6 A schematic diagram of the adjusted train control strategy for sections with insufficient power supply; Figure 7 This is the train timetable after the distributed adjustment across the entire line; Figure 8This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0014] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0015] This application can realize the coordinated control of trains in sections with insufficient power supply and the adjustment of train arrival and departure times in sections with sufficient power supply based on the information of traction power supply failure. By using automated and intelligent adjustment methods, the line capacity maintained during the failure period can be improved, and the adverse effects on train operation can be systematically reduced.

[0016] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] In one exemplary embodiment, a method for adjusting the operation of urban rail distributed trains is provided, which is applied to a scenario of insufficient unidirectional power supply; wherein, insufficient power supply means that the traction power provided by the traction power supply system is less than the operating power required by the urban rail train.

[0018] like Figure 1 As shown, the urban rail distributed train operation adjustment method includes: Step 100: Obtain information data from the urban rail distributed train system. This information data includes planned operating schedules, traction power supply fault information, and basic track and operational data for the urban rail trains.

[0019] Step 200: Construct a train cooperative control model for the power shortage section based on the information data. The train cooperative control model includes a first objective function and first constraints.

[0020] Step 300: The hybrid barrier deep Q-network algorithm is used to solve the train cooperative control model, determine the train cooperative control strategy for the power supply shortage section, adjust the train operation, and obtain the train operation adjustment result for the power supply shortage section.

[0021] Step 400: Based on the train operation adjustment results of the power supply shortage section, construct the train timetable adjustment model for the power supply shortage section; the train timetable adjustment model includes a second objective function and a second constraint condition; the power supply shortage ensures that the traction power provided by the traction power supply system is not less than the operating power required by the urban rail train.

[0022] Step 500: Solve the train timetable adjustment model using a preset solver, and combine it with the train cooperative control strategy for sections with insufficient power supply to obtain the adjusted train timetable. The train timetable includes: train arrival and departure times and operating routes.

[0023] In one embodiment, a hybrid barrier deep Q-network algorithm is used to solve the train cooperative control model to determine the train cooperative control strategy for sections with insufficient power supply, so as to adjust the train operation and obtain the train operation adjustment results for sections with insufficient power supply, specifically including: A distributed Markov decision framework is established, which forms a multi-agent system with all trains in the power-deficient section and observes the global state of the environment at each time step.

[0024] The joint action space of the agents is generated based on the global state, and the joint actions that satisfy the first constraint condition are determined to determine the train cooperative control strategy for the power shortage section. After the action is executed, the global state is updated, and the multi-agent system receives a reward. Based on the first objective function in the train cooperative control model, the discounted reward is maximized. The reward for each step is the total train running distance.

[0025] The process of determining the control strategy includes: Actions that violate the departure and departure-to-arrival interval constraints in the first constraint condition are removed using the PRS barrier, and then... - The strategy selects the original action, and checks whether the selected original action satisfies the power supply capacity constraint in the first constraint condition based on the POS barrier.

[0026] After selecting an action through the reselection mechanism, the transition experience with a penalty term is stored in the experience replay pool. A portion of the experience is sampled from the buffer of the experience replay pool and the control parameters of the deep Q network are updated using the stochastic gradient descent method to obtain the updated deep Q network. The joint action is determined based on the updated deep Q network and the global state. Each action within the joint action represents the control strategy of the train in the time step.

[0027] like Figure 2 As shown, in practical applications, the method described in this application includes the following steps: Step 1: Obtain the basic lines and operational data (basic lines and operational data), planned operation schedule information, and traction power supply fault information of the urban rail system.

[0028] Basic line and operational data include: a complete set of stations along the entire line, the shortest and longest stop times at each station, the shortest and longest travel times between adjacent stations, a set of depots with the capacity to store reserve trains, the number of reserve trains in each depot, the minimum departure interval, the minimum departure-arrival interval, and the shortest turnaround time for trains; planned operation schedule information includes: all trains in the planned operation schedule and sets of trains going up / down, the planned arrival / departure times of each train at each station in these two sets; traction power supply fault information includes: substations where DC feeder circuit breakers failed, fault duration, stations at both ends of the power shortage section, a set of stations in the power shortage section, the total traction power limit of the power shortage section, and the location and speed of all trains when the fault occurred.

[0029] Based on the actual situation of the urban rail transit system, configure the basic lines and operating parameters required for train operation adjustments: such as... Figure 3 As shown, the station set of the urban rail line is From station 1 to station E Defined as the up direction, the station E Station 1 is defined as the downstream direction, and the stations in this set are... The shortest and longest stopping times are respectively and From the station To the station The shortest and longest running times are respectively and There is one depot at each end of the line, denoted as... When the malfunction occurred, the depot... The quantity stored is The standby trains can be put into service to replace trains during a breakdown. The minimum departure interval is... The minimum departure-arrival interval is The shortest turnaround time for the train is .

[0030] Retrieve scheduled operation information: Set of all train services in the scheduled operation map. Upbound train collection Downbound train sets Train numbers in these two sets At the station Planned arrival time and planned departure time .

[0031] Obtain traction power supply fault information: located at the substation The DC feeder circuit breaker failed, and the fault duration was... The stations at both ends of the section with insufficient power supply are respectively and The stations in areas with insufficient power supply are grouped as follows: The total traction power in sections with insufficient power supply is limited to .

[0032] Step 2: With the goal of maximizing the average line capacity in the power-deficient section, and taking into account power supply capacity constraints and train operation constraints, establish a train cooperative control model to optimize the control strategy at each instant.

[0033] In step 2, the objective function of the model (i.e., the first objective function) is to maximize the average line capacity of the power-deficient section during the fault. The power supply capacity constraint limits the total traction power of the train at each instant. The train operation constraints include the maximum and minimum traction / braking force constraints, departure interval constraints, arrival and departure interval constraints, the longest and shortest stop time constraints, the longest and shortest inter-station running time constraints, and the line speed limit constraints.

[0034] To ensure that the line capacity in sections with insufficient power supply matches passenger demand as closely as possible, thereby mitigating the impact of traction power supply system failures on train operation, the objective function for train cooperative control is to maximize the average line capacity in sections with insufficient power supply during the failure period, as detailed below: The objective function (first objective function) for train cooperative control is: .

[0035] in, This represents the average line capacity. This represents the average departure interval; This represents the average number of trains. This is the average running time; This represents the average speed of the train. The length of the section with insufficient power supply; For train In time unit The running distance within; Total train running time; For train serial number; For trains assembled in areas with insufficient power supply; This is a set of time points during periods of insufficient power supply.

[0036] Defined as the number of trains passing a fixed point on the line within one hour, based on the average departure interval in sections with insufficient power supply. After direct calculation, the conversion is performed in three steps. First, the average departure interval is... This represents the average operating time within the power shortage area. With average number of trains The discussion then focused on the length of the power supply shortage section. Divided by the average speed of the train calculate .at last, It is expressed as the total distance the train traveled during the fault divided by the duration. In the form of. Among them, It is a train In time unit The running distance within the range is calculated using the following formula. It is a train station The location.

[0037] .

[0038] For train At the point of time The location.

[0039] In scenarios where unidirectional power supply is insufficient, only one substation supplies power to the entire area with insufficient power, which severely limits the total traction power of the train. Therefore, power supply capacity constraints need to be established to avoid overload, as detailed below: First, calculate the instantaneous power of each train. The formula for calculating the mechanical power at the train axle is: ,in and They are trains At any moment The applied force (traction or braking force) and speed. Considering power losses in processes such as transformer, inverter, and gear transmission, the energy utilization efficiency is set to... Furthermore, the auxiliary power of onboard auxiliary equipment (such as air conditioning and lighting equipment) is assumed to be a constant value. Therefore, the traction power at each moment can be calculated using the following formula: .

[0040] Due to limited traction and braking performance, the applied force The following constraints must be met: .

[0041] in, and They are speeds The maximum traction force and maximum braking force under the given conditions, and the expressions corresponding to the traction / braking force constraints are: .

[0042] .

[0043] The traction and braking curves are roughly divided into two sections. In the first section, as the train speed increases, the traction force... and braking force Maintain a constant speed. Once the speed exceeds the threshold... and Maximum traction and braking force will decrease with speed. Assuming the decrease after exceeding the threshold is linear, the rates of decrease are respectively... and .

[0044] In addition, calculating instantaneous train power requires obtaining speed. and location The calculation formula is as follows: .

[0045] in, and These are the basic resistance and the additional resistance. Basic resistance... Due to the influence of various factors such as bearing resistance, rolling resistance, sliding resistance, and air friction, precise calculation is difficult. Therefore, the Davis equation is used for approximate calculation, i.e. ,in Davis's coefficient is related to friction between wheels and rails, as well as train aerodynamics. Additional drag. Depending on the environmental conditions, the formula is used. Approximate calculation of additional drag, where It is a location The slope angle at that location.

[0046] The instantaneous power of a train may vary at different times within a time unit. To avoid overload, the maximum traction power within each time unit is used to represent the train's traction power, calculated as follows: .

[0047] Finally, the following power supply capacity constraints are established to control the total traction power within the rated range. This refers to the total traction power limit in sections with insufficient power supply, i.e., the maximum total traction power in sections with insufficient power supply. The expression corresponding to the power supply capacity constraint is: .

[0048] in, For train In time unit The train traction power within the system; For train serial number; This represents the maximum total traction power in sections with insufficient power supply. For trains assembled in areas with insufficient power supply; This is a set of time points during periods of insufficient power supply. Index of time points during periods of power shortage; Index for the final time point; For a specific moment; For a moment traction power; For train At any moment The applied force; For train At any moment and speed; For the efficiency of electrical energy utilization; This refers to the auxiliary power of onboard auxiliary equipment.

[0049] To ensure the feasibility of train operation adjustment schemes in scenarios of insufficient one-way power supply, train operation constraints are established for sections with insufficient power supply, including maximum and minimum traction / braking force constraints, departure interval constraints, arrival and departure interval constraints, longest and shortest stop time constraints, longest and shortest inter-station travel time constraints, and line speed limit constraints, as detailed below: To ensure operational safety, the following departure interval constraints guarantee that the time interval between consecutive departures from the same station is not less than [amount missing]. Meanwhile, to avoid severely unbalanced departure intervals affecting service quality, the interval should not exceed [a certain value]. The expression corresponding to the departure interval constraint is: .

[0050] The departure time is determined based on the following three conditions. : like , , ,but . This refers to the number of train operation cycles in sections with insufficient power supply. For train In time unit The distance traveled within the area.

[0051] Similarly, the interval between the departure of the preceding vehicle and the arrival of the following vehicle is limited to being greater than or equal to [a certain value]. The expression corresponding to the arrival / departure interval constraint is: .

[0052] The following stop time constraint allows sufficient passenger boarding and alighting time while maintaining service quality. That is, the expression corresponding to the stop time constraint is: .

[0053] Inter-station travel time constraints are related to factors such as track conditions, traction / braking performance, and passenger service. The expression corresponding to the inter-station travel time constraint is: .

[0054] Line speed limits are used to prevent trains from speeding or stopping between stations, among which It is a location The speed limit at this location. The expression corresponding to the speed limit constraint is: .

[0055] in, For speed Maximum traction force; For traction force; The velocity threshold under traction force; The rate of descent under traction force; For speed The maximum braking force under the current; For braking force; The speed threshold under braking force; The rate of descent under braking force; This is the minimum time interval between departures; This is the maximum time interval between departures; For train At the station Departure time; For train At the station Departure time; A collection of stations in areas with insufficient power supply; For trains assembled in areas with insufficient power supply; This is the minimum time interval between departure and arrival. For train At the station Arrival time; One of the stations at either end of the section with insufficient power supply; Minimum stop time; For train At the station Arrival time; This is the maximum stop time; For train At the station Arrival time; Minimum inter-station travel time; This is the maximum inter-station travel time; For train At the point of time speed; For position Speed ​​limits are in place at this location; For train At the point of time Location; This is a set of station locations in areas with insufficient power supply.

[0056] Step 3: Based on the results of train operation adjustments in sections with insufficient power supply, and with the goal of maximizing the average line capacity in sections with sufficient power supply, a train timetable adjustment model is established to optimize train arrival and departure times and routes by combining various scheduling measures such as adjusting stop times, adjusting inter-station travel times, and canceling train services.

[0057] In step 3, the objective function of the model (the second objective function) is to maximize the average line capacity of the section with sufficient power supply during the fault and minimize the unmet passenger demand. Considering five scheduling measures, namely, adjusting the stop time, adjusting the inter-station running time, canceling trains, train entering the depot, and putting reserve trains into service, the constructed train operation constraints (i.e., the second constraints) include cross-section time coupling constraints, train entering and leaving the depot constraints, arrival and departure time constraints, departure / arrival and departure interval constraints, and terminal station turnaround constraints.

[0058] Based on the results of train operation adjustments in sections with insufficient power supply, and with the goal of maximizing the average line capacity in sections with sufficient power supply and minimizing the number of passengers not being served, a train timetable adjustment model is established to optimize train arrival and departure times and routes by combining various scheduling measures such as adjusting stop times, adjusting inter-station travel times, and canceling train services.

[0059] In adjusting train schedules in sections with sufficient power supply, the negative impact of reduced capacity in sections with insufficient power supply on the capacity in sections with sufficient power supply is considered, as well as passenger demand, systematically reducing the impact on passenger service. Therefore, the objective function in the model is to maximize the average line capacity in sections with sufficient power supply during the fault period and minimize the unmet passenger demand, as detailed below: Objective function (second objective function) for sections with sufficient power supply It consists of two parts. It is a weighting coefficient that adjusts the two parts to the same order of magnitude.

[0060] .

[0061] Since the more stops a line makes within a given time period, the higher its capacity is generally, therefore the first part of the second objective function... The aim is to maximize the number of train stops during outages by focusing on operations, thereby increasing the capacity maintained in sections with sufficient power supply. .

[0062] in, It is to determine the train number Were you at the station during the malfunction? The binary variable of parking.

[0063] .

[0064] in, and All are binary variables. Used to determine train number Arrival at the station Whether the time is before the fault ends can be indicated by the following IF-THEN rule: .

[0065] Arrival at the station Time; It is the end time of the traction power supply failure.

[0066] In addition, binary variables Used to determine train number At the station Will the planned stops be cancelled? If there are no backup or turnaround trains operating, then the train service will proceed. If so, the train service can only be cancelled. ;otherwise, .therefore, The calculation is expressed as: .

[0067] Among them, binary variables Determine train number Is it from the depot? The train was ready to depart. Used to determine the number of trains operating. Will the train continue operating after turning back? .

[0068] The second term of the second objective function represents the number of passenger demands for service that are not being met. The calculation is as follows: .

[0069] in, Even after transportation services were provided during the outage, the station and Unmet passenger demand remains. Based on time-of-day passenger demand and the number of service trains in each section during the outage, the sections with sufficient power supply... The corresponding expression is: .

[0070] in, Indicates the station during the malfunction and Total passenger demand between This refers to the rated passenger capacity of each train trip, and the number of service trains per section is determined by... calculate.

[0071] Considering five scheduling measures—adjustment of stop times, adjustment of inter-station travel times, cancellation of train services, train entry into the depot, and preparation of reserve trains—the constructed train operation constraints, i.e., the second constraint conditions, include cross-segment time coupling constraints, train entry and exit constraints, arrival and departure time constraints, departure interval constraints, arrival and departure interval constraints, and terminal station turnaround constraints, as detailed below: Cross-segment time coupling constraints: like Figure 4 As shown, in order to ensure the feasibility of the train operation adjustment plan for the entire line, it is necessary to construct cross-segment time coupling constraints under different scenarios based on the initial position of the train.

[0072] Scenario 1: If the train malfunctions at the time... If the train is located in a section with insufficient power supply, it will continue to run its scheduled route during the fault. ,Right now Among them, binary variables Used to determine train Is there a train service? In this scenario, the train's departure time at the boundary between sections with insufficient and sufficient power supply... The following cross-segment time coupling constraints should be met.

[0073] .

[0074] Scenario 2: If the train malfunctions at the time... If the train is located outside the area with insufficient power supply, the train service will be uncertain during the fault. If Then the train number The departure time at the boundary should be consistent with the train's departure time in the collaborative control results. The timing is consistent.

[0075] .

[0076] in, For trains in the collaborative control results At the station Arrival time; For train At the station The actual arrival time; This refers to the group of trains located outside the section with insufficient power supply when the fault occurs.

[0077] To ensure time coupling across different sections, some additional constraints must be met. Among these, each train should operate one trip within sections with insufficient power supply. .

[0078] Secondly, each train service must either be cancelled or operated by a single train, as indicated by: .

[0079] Finally, there should be enough trains to operate each trip in sections with insufficient power supply, as follows: .

[0080] Among them, binary variables Determine the train number During the malfunction, was the station... Departure is indicated as: .

[0081] in, The end time of the traction power supply failure.

[0082] Train entry and exit constraints: Prior to the malfunction, the train entry and exit procedures were as planned, meaning there were no trains entering or leaving the depot. .

[0083] in, and They are train numbers At the station The planned arrival and departure times. If the train is operating on a certain schedule... Then return to the depot ,but ;otherwise, Furthermore, if the depot is able to resume service through standby train measures... The backup vehicle entered the main line to replace the train. ,but ;otherwise, .

[0084] In addition, the number of reserve cars in each depot is limited, which restricts the total number of trains that can leave the depot and enter service. .

[0085] Among them, the depot The number of spare vehicles stored is .

[0086] Arrival and departure time constraints: Prior to the malfunction, the train was operating according to the scheduled timetable. Therefore, the actual arrival and departure times should be consistent with the scheduled times. .

[0087] The station dwell time constraints and inter-station travel time constraints are as follows: .

[0088] .

[0089] Departure / arrival interval constraints: To ensure safe train operation, the time interval between any two train departures must meet the following requirements: .

[0090] Among them, binary variables Determine train number and Were they all not at the station? Cancel. If so, the departure interval should not be less than [amount missing]. and . The calculation formula is: .

[0091] In addition, the minimum time interval between the departure of the preceding vehicle and the arrival of the following vehicle should not be less than [amount missing]. : .

[0092] Terminal station turnaround constraints: Upon arrival at the terminal station, the train will turn back and operate in the opposite direction. Due to factors such as track infrastructure and signaling systems, the turnaround time should not be less than the minimum turnaround time. : .

[0093] Among them, binary variables Determine the number of trains to run Will the train turn back and resume service? .

[0094] In addition, the train connection relationship should meet the following three constraints: each train cannot be operated by a standby train and a turnaround train at the same time; after a train departs from the terminal station, it cannot simultaneously enter the depot and turn around; if a train does not enter the depot after arriving at the terminal station, it needs to turn around and operate the next train.

[0095] .

[0096] .

[0097] .

[0098] Step 4: Use reinforcement learning algorithm and GUROBI solver to solve the train operation adjustment problem in sections with insufficient power supply and sufficient power supply in real time, and systematically obtain the adjustment plan for the entire line.

[0099] Step 4, based on the characteristics that power supply capacity constraints, departure interval constraints, and arrival / departure interval constraints in the train cooperative control model for power-deficient sections in Step 2 are all closely related to train operation safety, incorporates a protection mechanism into the classic DQN algorithm, designing a Hybrid Barrier Deep Q-Network (HSDQN) algorithm. This safety reinforcement learning algorithm additionally introduces pre- and post-barriers to prevent the execution of unsafe actions that violate safety constraints. Furthermore, a penalty term is added to the transfer experience involving unsafe actions to reduce the probability of selecting these actions again. Using the above method, an offline train cooperative control decision-maker for unidirectional power shortage scenarios is trained. The basic operating parameters and traction power supply fault information obtained in Step 1 are input into the decision-maker, which outputs a train operation adjustment scheme for power-deficient sections. Based on this, the linearized train timetable adjustment model for power-sufficient sections in Step 4 is solved using GUROBI and other methods to obtain the train operation adjustment scheme for power-sufficient sections.

[0100] A Hybrid-Shield Deep Q-Network (HSDQN) algorithm is designed to solve the train cooperative control problem in section 2. This algorithm first constructs a distributed Markov decision process framework. For example... Figure 5 As shown, all trains in the section with insufficient power supply This constitutes a multi-agent system, with the central controller at each time step. Global state of the observation environment , It consists of three parts, including the combined status of all trains' speed, position, and time spent stopping / traveling. The joint state formed by the elapsed time since the last departure of each station. and remaining time of failure Subsequently, a joint action space for the agent is generated based on the observed state. According to deterministic strategies Determine the joint actions that satisfy the departure interval constraints. Each action represents the train's control strategy within a time step. After an action is executed, the environment state changes according to the function. from Updated to Meanwhile, the multi-agent system receives a reward. To maximize the line capacity in areas with insufficient power supply, the reward for each step is defined as the total distance the train travels. The goal of this algorithm is to maximize the discount return. ,in, This is the discount factor that ensures convergence. If an unsafe action is chosen, the following out-of-limit costs will be penalized: .

[0101] To ensure driving safety, the algorithm employs three protection mechanisms to address various safety-related constraints: (1) First, the compressed state space method is used to handle safety-related constraints such as speed limits, dwell time, and running time. Specifically, if or This status will be removed due to violation of the stop time constraint. Similarly, if or This status will be removed because the inter-station travel time constraint is not met. If the train speed... Lines that do not meet speed limit constraints will also be removed.

[0102] (2) Secondly, two safety barriers are set up to prevent unsafe actions that violate the constraints of departure interval and power supply capacity: 1) Set up a preliminary safety barrier to address departure interval constraints. PRS At each time step, the barrier will calculate a set of safety actions based on the environmental state. All violations of departure interval constraints (when or From the station Departure) and arrival-departure interval constraints (when Arrive at the station Unsafe actions will be removed. This method ensures that only actions that satisfy departure or arrival-departure interval constraints are available.

[0103] 2) Set up a backup safety barrier to address power supply capacity constraints. POS At each time step, the agent initially moves from the action space. Choose any action. If an unsafe action that would cause an overload is selected, a safety barrier will prevent its execution and select another safe action instead. This method significantly reduces the computational complexity required to verify whether power supply capacity constraints are met.

[0104] (3) Finally, a penalty is set to reduce the probability of choosing unsafe actions during training: if overload is caused by taking unsafe actions, an over-limit penalty is imposed. The rewards for this state-action pair will be included. Subsequently, the transferred experience with penalties will be added to the experience replay pool, significantly reducing the likelihood of choosing actions that lead to overload.

[0105] By combining the aforementioned protection mechanisms with the classic DQN algorithm, a safety reinforcement learning algorithm, HSDQN, is designed to solve the train cooperative control problem. In each step, the algorithm first... PRS Barrier removal eliminates unsafe actions that violate departure and departure-arrival interval constraints. Then, using... - The strategy selects the original action. POS The barrier check ensures that the selected action meets power supply capacity constraints. After a safe action is selected through the reselection mechanism, the transferred experience is stored in the experience replay pool. During training, mini-batch experiences are sampled from the buffer, and the neural network controller parameters are updated using stochastic gradient descent to achieve controller pre-training. In actual operation, when unidirectional power supply insufficiency occurs, the environmental state can be input into the pre-trained controller in real time to calculate the train control strategy for the power supply insufficient section, such as... Figure 6 As shown.

[0106] The train timetable adjustment model for sections with sufficient power supply is solved using solvers such as GUROBI within a specified time. The adjusted train timetable is then obtained by combining the train cooperative control strategy for sections with insufficient power supply. Figure 7 This diagram shows the adjusted train schedule for a certain section of a subway line in a city experiencing a 1500-second one-way power shortage in the northbound direction starting at 8:00 AM. Gray shading indicates temporary train decommissioning during periods of one-way power supply, red shading indicates train operation adjustments for sections with insufficient power, and yellow shading indicates train operation adjustments for sections with sufficient power.

[0107] This application acquires basic line operation data, planned timetable information, and traction power supply fault information from the urban rail transit system. This data is then input into train operation adjustment models for sections with insufficient and sufficient power supply. The adjusted train operation adjustment schemes are solved using a safety reinforcement learning algorithm and GUROBI, respectively, achieving automated and intelligent distributed train operation adjustment. This reduces the impact of traction power supply faults on train operation and improves the actual timetable fulfillment rate, adapting to the refined management of the urban rail transit system.

[0108] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 8As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores urban rail distributed train operation adjustment data. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements the urban rail distributed train operation adjustment method.

[0109] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0110] In one exemplary embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0111] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0112] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0113] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0114] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0115] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, data processing logic devices, etc., and are not limited to these.

[0116] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0117] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for adjusting the operation of distributed trains in urban rail transit, characterized in that, The urban rail distributed train operation adjustment method is applied to scenarios with insufficient one-way power supply; where insufficient power supply means that the traction power provided by the traction power supply system is less than the operating power required by the urban rail train. The method for adjusting the operation of distributed trains on urban rail transit includes: Acquire information data of the urban rail distributed train system; the information data includes planned operation schedule information, traction power supply fault information, and basic line and operation data of urban rail trains; A train cooperative control model for the power shortage section is constructed based on the information data; the train cooperative control model includes a first objective function and a first constraint condition; The hybrid barrier deep Q-network algorithm is used to solve the train cooperative control model to determine the train cooperative control strategy for the power shortage section, so as to adjust the train operation and obtain the train operation adjustment result for the power shortage section. Based on the train operation adjustment results for sections with insufficient power supply, a train timetable adjustment model for sections with sufficient power supply is constructed. The train timetable adjustment model includes a second objective function and a second constraint condition. Sufficient power supply means that the traction power provided by the traction power supply system is not less than the operating power required by the urban rail train. The train timetable adjustment model is solved using a preset solver, and combined with the train cooperative control strategy for sections with insufficient power supply, to obtain the adjusted train timetable; the train timetable includes: train arrival and departure times and operating routes.

2. The urban rail distributed train operation adjustment method according to claim 1, characterized in that, The first objective function is constructed with the goal of maximizing the average line capacity in sections with insufficient power supply during traction power supply failures; the expression of the first objective function is: ； in, This represents the average line capacity. This represents the average departure interval; This represents the average number of trains. This is the average running time; This represents the average speed of the train. The length of the section with insufficient power supply; For train In time unit The running distance within; Total train running time; For train serial number; For trains assembled in areas with insufficient power supply; This is a set of time points during periods of power shortage. Index of time points during periods of power shortage; This is the index for the final time point.

3. The urban rail distributed train operation adjustment method according to claim 1, characterized in that, The first constraint includes: power supply capacity constraint and train operation constraint; The power supply capacity constraint limits the total traction power of the train at any instant; the expression corresponding to the power supply capacity constraint is: ； ； ； in, For train In time unit The train traction power within the system; For train serial number; This represents the maximum total traction power in sections with insufficient power supply. For trains assembled in areas with insufficient power supply; This is a set of time points during periods of power shortage. Index of time points during periods of power shortage; Index for the final time point; For a specific moment; For a moment traction power; For train At any moment The applied force; For train At any moment and speed; For the efficiency of electrical energy utilization; Auxiliary power for vehicle-mounted auxiliary equipment; The train operation constraints include traction / braking force constraints, departure interval constraints, arrival and departure interval constraints, station dwell time constraints, inter-station travel time constraints, and line speed limit constraints. The expression corresponding to the traction / braking force constraint is: ； ； The expression corresponding to the departure interval constraint is: ； The expression corresponding to the arrival / departure interval constraint is: ； The expression corresponding to the stop time constraint is: ； The expression corresponding to the inter-station running time constraint is: ； The expression corresponding to the line speed limit constraint is: ； in, For speed Maximum traction force; For traction force; The velocity threshold under traction force; The rate of descent under traction force; For speed The maximum braking force under the current; For braking force; The speed threshold under braking force; The rate of descent under braking force; This is the minimum time interval between departures; This is the maximum time interval between departures; For train At the station Departure time; For train At the station Departure time; A collection of stations in areas with insufficient power supply; This represents the average number of trains. For trains assembled in areas with insufficient power supply; This is the minimum time interval between departure and arrival. For train At the station Arrival time; One of the stations at either end of the section with insufficient power supply; Minimum stop time; For train At the station Arrival time; This is the maximum stop time; For train At the station Arrival time; Minimum inter-station travel time; This is the maximum inter-station travel time; For train At the point of time speed; For train At the point of time Location; For position Speed ​​limits are in place at this location; This is a set of station locations in areas with insufficient power supply.

4. The urban rail distributed train operation adjustment method according to claim 1, characterized in that, The second objective function is determined by maximizing the average line capacity in areas with sufficient power supply and minimizing the number of unmet passengers; the expression for the second objective function is: ； in, The second objective function; This refers to the number of times the train stopped during the malfunction. The number of passengers whose needs are not being met; These are the weighting coefficients.

5. The urban rail distributed train operation adjustment method according to claim 1, characterized in that, The second constraint includes cross-segment time coupling constraint, train entry and exit constraint, arrival and departure time constraint, departure interval constraint, arrival and departure interval constraint, and terminal station turnaround constraint.

6. The urban rail distributed train operation adjustment method according to claim 1, characterized in that, The hybrid barrier deep Q-network algorithm is used to solve the train cooperative control model to determine the train cooperative control strategy for sections with insufficient power supply, so as to adjust the train operation and obtain the train operation adjustment results for sections with insufficient power supply, specifically including: A distributed Markov decision framework is established, which forms a multi-agent system with all trains in the power-deficient section and observes the global state of the environment at each time step. The joint action space of the agents is generated based on the global state, and the joint actions that satisfy the first constraint condition are determined to determine the train cooperative control strategy for the power shortage section. After the action is executed, the global state is updated, and the multi-agent system receives a reward. Based on the first objective function in the train cooperative control model, the discounted reward is maximized. The reward for each step is the total train running distance. The process of determining the control strategy includes: Actions that violate the departure and departure-to-arrival interval constraints in the first constraint condition are removed using the PRS barrier, and... - The strategy selects the original action, and checks whether the selected original action satisfies the power supply capacity constraint in the first constraint condition based on the POS barrier. After the action is selected through the reselection mechanism, the transition experience with the penalty term will be stored in the experience replay pool. A portion of the experience is sampled from the buffer of the experience replay pool and the control parameters of the deep Q network are updated using the stochastic gradient descent method to obtain the updated deep Q network. The joint action is determined based on the updated deep Q-network and the global state; each action within the joint action represents the control strategy of the train within a time step.

7. The urban rail distributed train operation adjustment method according to claim 6, characterized in that, The expression corresponding to the penalty term is: ； in, This is a penalty item; This represents the maximum total traction power in sections with insufficient power supply. For train In time unit Internal traction power; For train In time unit The internal renewable energy utilization power.

8. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the urban rail distributed train operation adjustment method according to any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the urban rail distributed train operation adjustment method as described in any one of claims 1-7.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the urban rail distributed train operation adjustment method as described in any one of claims 1-7.

Images