A virtual marshalling train energy-saving optimization method, device and product

Abstract

The application discloses a virtual marshalling train energy-saving optimization method, device and product, relates to the technical field of virtual marshalling train energy-saving, and comprises the following steps: acquiring line data and train historical operation data of different types of trains, and calculating train operation energy consumption of the different types of trains; based on a virtual marshalling technology and a mixed integer programming method, a train diagram model is constructed according to the train operation energy consumption; the train diagram model comprises constraint conditions and a target function, the constraint conditions comprise train timetable constraints, passenger flow constraints and train coupling constraints, and the target function aims to balance passenger waiting time and train operation energy consumption; the train diagram model is solved to obtain a train diagram; and according to the train historical operation data, the train diagram and the line data, an optimal train operation speed curve is generated by using a double-depth Q network algorithm. Through the virtual marshalling technology, the mixed integer programming method and reinforcement learning, the optimization of train energy consumption is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of virtual marshalling train energy saving, in particular to a virtual marshalling train energy saving optimization method, device and product. BACKGROUND

[0002] Rail transit transportation system is the basis for the survival and function of the entire city system. Urban rail transit is increasingly valued by countries around the world due to its advantages of fast speed, large capacity, safety and energy saving. At present, the urban rail transit network in China is developing rapidly. With the rapid expansion of the urban rail transit network, the problem of energy consumption has become increasingly serious and has become a key challenge that needs to be addressed urgently. Therefore, how to effectively improve energy use efficiency and reduce energy consumption has become a key issue for the optimization of rail transit transportation system, which is of great significance for the sustainable development of cities and the realization of carbon emission peak target.

[0003] There are many factors that affect train operation energy consumption, but they can be generally divided into three categories: traction energy consumption, auxiliary energy consumption and station energy consumption. In practical applications, traction energy consumption occupies a dominant position in train operation energy consumption. In view of the key role of traction energy consumption in train operation energy consumption and its operability, the purpose is to adjust the train operation strategy by optimizing the traction energy consumption to achieve energy saving effect.

[0004] In order to meet the demand of train energy saving, there are usually two solutions: one is to optimize the train running speed curve, and the other is to optimize the train running schedule. Both of them are hotspots in the rail transit industry in recent years. Although the above research provides good theoretical support for the energy-saving operation of urban rail transit, but at present these researches are based on fixed marshalling, that is, only large trains or small trains are used to run the line. In recent years, with the rapid development of intelligent technology, virtual marshalling technology has been gradually introduced into the field of rail transit. Virtual marshalling technology can use large trains and small trains to run the line at the same time, which is one of the key means to improve train operation efficiency and line capacity. On the basis of virtual marshalling technology, how to further optimize the operation energy consumption of trains, reduce energy consumption and improve energy efficiency has become an important problem in the current research of rail transit field. SUMMARY

[0005] The purpose of the present application is to provide a virtual marshalling train energy saving optimization method, device and product, which can optimize the energy consumption of trains by using virtual marshalling technology, mixed integer programming method and reinforcement learning.

[0006] To achieve the above purpose, the present application provides the following solutions:

[0007] In a first aspect, the present application provides a virtual marshalling train energy saving optimization method, comprising:

[0008] acquire line data and train historical operation data of different types of trains, and calculate train operation energy consumption of the different types of trains according to the line data and the train historical operation data; the line data comprises distances between stations, slopes and train speed limits, and the train historical operation data comprises positions, speeds, accelerations and operation resistances of the trains at current time points;

[0009] based on a virtual marshalling technology and a mixed integer programming method, a train diagram model is constructed according to the train operation energy consumption; the train diagram model comprises constraint conditions and a target function, the constraint conditions comprise train timetable constraints, passenger flow constraints and train coupling constraints, and the target function is a function aiming to balance passenger waiting time and the train operation energy consumption;

[0010] the train diagram model is solved to obtain a train diagram;

[0011] according to the train historical operation data, the train diagram and the line data, a double deep Q network algorithm is used to generate an optimal train operation speed curve.

[0012] In a second aspect, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor executes the computer program to implement the virtual marshalling train energy-saving optimization method in the above.

[0013] In a third aspect, a computer program product is provided, comprising a computer program executable by a processor to implement the virtual marshalling train energy-saving optimization method in the above.

[0014] According to the specific embodiments provided in the present application, the following technical effects are disclosed:

[0015] The present application provides a virtual marshalling train energy-saving optimization method, device and product, and the mixed integer programming method is used to optimize the train diagram, so that the waiting time of passengers and the energy consumption of trains can be balanced while the timetable constraints are guaranteed. Moreover, the double deep Q network algorithm is used to optimize the train operation speed curve according to the train historical operation data and the train diagram, so as to minimize the energy consumption and meet other constraint conditions. Reinforcement learning can process dynamic changes in the environment, so as to ensure that the train operation speed is always optimal. It can be seen that the virtual marshalling technology, the mixed integer programming method and the deep reinforcement learning are comprehensively used to realize accurate calculation and optimization of train energy consumption. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings described below only illustrate some of the embodiments of the present application, and for those skilled in the art, other drawings can be obtained from these drawings without creative labor.

[0017] Figure 1 An application environment diagram of a virtual marshalling train energy-saving optimization method in an embodiment of the present application;

[0018] Figure 2 A flowchart of a virtual marshalling train energy-saving optimization method provided in an embodiment of the present application;

[0019] Figure 3 A functional module diagram of a double-depth Q network;

[0020] Figure 4 A diagram of an optimal train running speed curve;

[0021] Figure 5 A performance diagram of fixed marshalling coupling and virtual marshalling coupling strategies with different weight ratios;

[0022] Figure 6a A passenger waiting time performance diagram of fixed marshalling coupling and virtual marshalling coupling strategies in different instances under non-peak hours;

[0023] Figure 6b A train energy consumption diagram of fixed marshalling coupling and virtual marshalling coupling strategies in different instances under non-peak hours;

[0024] Figure 7a A passenger waiting time performance diagram of fixed marshalling coupling and virtual marshalling coupling strategies in different instances under peak hours;

[0025] Figure 7b A train energy consumption diagram of fixed marshalling coupling and virtual marshalling coupling strategies in different instances under peak hours;

[0026] Figure 8 A structural diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0028] Train diagram model contains three parts of train, passenger flow and energy consumption. Train diagram optimization needs to consider not only basic conditions such as station, line and train bottom resources, but also multiple constraints such as marshalling operation, energy consumption limit, safety interval and dynamic passenger flow. The introduction of virtual marshalling strategy is likely to bring nonlinear constraints and expand the scale of constraints, which will bring great challenges to model solving. These complex constraints are often intertwined, and how to balance between these conditions when optimizing is a major challenge in model solving, which not only guarantees efficient train operation, but also ensures passenger experience and system safety. In addition, how to maintain the accuracy of the model while ensuring the calculation efficiency is also one of the difficulties to be overcome.

[0029] The passenger flow of urban rail transit has significant dynamic spatial and temporal characteristics, and passenger flow fluctuation has a direct impact on train scheduling and marshalling decision. In actual operation, the passenger flow of different time periods and different stations varies greatly. How to dynamically respond to passenger flow demand and reasonably adjust the marshalling form of the train in the train diagram optimization is also one of the core difficulties of model optimization. At the same time, the uncertainty of passenger flow increases the complexity of the model, which requires quick response in optimization and ensures the stability and robustness of the model.

[0030] In order to make the above purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0031] The virtual marshalling train energy-saving optimization method provided by the embodiments of the present application can be applied to, for example Figure 1The application environment is shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data required by the server 104 to process. The data storage system can be set up separately, or integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the line data and the train historical operation data of different types of trains to the server 104, and the server 104 calculates the train operation energy consumption of different types of trains according to the line data and the train historical operation data of different types of trains; based on the virtual marshalling technology and the mixed integer programming method, according to the train operation energy consumption, a train diagram model is constructed; the train diagram model is solved to obtain a train diagram; then a double deep Q network algorithm is used to generate an optimal train running speed curve based on the train historical operation data, the train diagram and the line data. The server 104 can feed back the optimal train running speed curve obtained to the terminal 102. In addition, in some embodiments, the virtual marshalling train energy saving optimization method can also be implemented by the server 104 or the terminal 102 alone, such as can be directly processed by the terminal 102 for line data and train historical operation data of different types of trains, or the server 104 can obtain line data and train historical operation data of different types of trains from the data storage system and process them.

[0032] Among them, the terminal 102 can be but not limited to various desktop computers, notebook computers and Internet of Things devices portable wearable devices. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers, and can also be a cloud server.

[0033] In an exemplary embodiment, as Figure 2 shown, a virtual marshalling train energy saving optimization method is provided, which is executed by a computer device, specifically can be executed by a terminal or a server computer device alone, or can be executed by a terminal and a server together. In the embodiment of the present application, the method is applied to the server 104 in Figure 1 The following steps 201 to 204 are described. Among them:

[0034] Step 201, obtaining line data and train historical operation data of different types of trains, and calculating train operation energy consumption of different types of trains according to line data and train historical operation data of different types of trains; the line data includes the distance between each station, the slope and the train speed limit, and the train historical operation data includes the current time position, speed, acceleration and running resistance of the train.

[0035] The wheel-rail force sensor, the traction force sensor and the aerodynamic resistance sensor integrated on the train can calculate the running resistance of the train under different speeds and conditions. In addition, the acceleration sensor, the speed sensor and the positioning device installed on the train can collect the acceleration, speed and position data sequence of the train. After defining the time interval as a specific value, the train historical running data table of the train formation within M time steps can be obtained as follows:

[0036] Table 1. Train historical running data table

[0037]

[0038] In step 202, based on the virtual marshalling technology and the mixed integer programming method (MIP), a train diagram model is constructed according to the train running energy consumption; the train diagram model includes constraint conditions and an objective function, the constraint conditions include train timetable constraints, passenger flow constraints and train coupling constraints, and the objective function is a function of balancing passenger waiting time and train running energy consumption.

[0039] In step 203, the train diagram model is solved to obtain a train diagram.

[0040] In step 204, according to the train historical running data, the train diagram and the line data, an optimal train running speed curve is generated by using a double deep Q network (DQN) algorithm.

[0041] By implementing the above steps 201 to 204, the train diagram is optimized by the mixed integer programming method, so that the passenger waiting time and the energy consumption of the train can be balanced while ensuring the timetable constraints. Moreover, by using the double deep Q network algorithm, the train running speed curve is optimized according to the train historical running data and the train diagram to minimize the energy consumption and meet other constraint conditions. The reinforcement learning can process the dynamically changing environment to ensure that the train running speed is always optimal. It can be seen that by comprehensively using the virtual marshalling technology, the mixed integer programming method and the deep reinforcement learning, the precise calculation and optimization of the train energy consumption are realized.

[0042] Further, the calculation of the train running energy consumption of different types of trains according to the line data and the train historical running data of different types of trains involved in step 201 is described in detail. Specifically, the train running energy consumption of different types of trains is calculated according to the following formula:

[0043]

[0044] wherein v i (0,ξ m ) is the mass of the train at station i with mass ξm The M-type train at t=0 speed; For the time spent at the station When it arrives at station i+1, its mass is ξ m The speed of the M-type train; v i (t,ξ m Let t be the speed of the train at station i at time t; Let ξ be the travel time of train k at station i; m The mass of train model m; V i min (t) represents the minimum speed limit at station i at time t; S represents the set of stations; M represents the set of virtual train types; and K represents the set of train numbers.

[0045]

[0046] Where v(t+1) is the speed of the train at time t+1; x(t+1) is the distance the train travels at time t+1; v(t) is the speed of the train at time t; x(t) is the distance the train travels at time t; λ B (t,ξ m ) represents the ratio of the braking force output by the train per unit time; λ F (t,ξ m B[v] represents the ratio of the traction force output by the train per unit time; i (t,ξ m Let v be the speed of train m at station i. i (t,ξ m Braking force under ); r[v i (t,ξ m Let v be the speed of train m at station i. i (t,ξ m The basic resistance under (g) i (t) represents the slope of station i at time t, i.e., the gradient force; F[v i (t,ξ m Let v be the speed of train m at station i. i (t,ξ m The traction force under )

[0047]

[0048] r(v) = a1 + a2v + a3v 2 ;

[0049] Where F(v) is the traction force when the train's speed is v; and B(v) is the braking force when the train's speed is v. Maximum traction force; is the maximum traction power; v z is the breaking speed limit; is the maximum braking force; is the maximum braking power; v b is the breaking speed limit; r(v) is the basic resistance of the train at speed v, including frictional resistance and air resistance; a1, a2, a3 are Davis equation coefficients representing the train parameter characteristics.

[0050]

[0051] wherein E m is the total energy consumption of the train, composed of the cumulative traction energy consumption generated between each station; x is the running distance of the train between stations.

[0052] Further, the train schedule constraints in step 202 include train section running time constraints and train arrival and departure time constraints.

[0053] The train section running time constraints are as follows:

[0054]

[0055] wherein, is the departure time of train k+1 at station i; is the departure time of train k at station i; T min is the safe departure time between consecutive trains.

[0056] The train arrival and departure time constraints are as follows:

[0057]

[0058] wherein, is the arrival time of train k at station i; is the departure time of train k at station i-1; is the running time of train k from station i-1.

[0059] Further, the train coupling constraints in step 202 are as follows:

[0060]

[0061] wherein y k,m is the decision variable, ensuring that the train can only choose one marshalling type.

[0062] Further, the passenger flow constraints in step 202 include the number of passengers getting off, the number of waiting passengers, the remaining passenger capacity, the number of passengers getting on, and the number of passengers on the train.

[0063] The number of passengers getting off at the station is:

[0064]

[0065] wherein pa k,i is the number of passengers getting off train k at station i, pr k,i-1 is the number of remaining passengers who did not get on train k at station i-1; A i is the ratio.

[0066] The number of waiting passengers is:

[0067]

[0068] wherein pw k,i is the number of passengers waiting for train k at station i; pw k-1,i is the number of passengers who failed to get on the previous train k-1 at station i; pb k-1,i is the number of passengers who arrived at station i after the departure of the previous train k-1; is the off-boarding rate of remaining passengers in the train at station i at time t; is a slack variable, which is the offset of the departure time of train k from station i relative to t when train k departs at [t, t+1], and is 0 when train k does not depart; is a slack variable, which is the offset of the departure time of train k-1 from station i relative to t when train k-1 departs at [t, t+1], and is 0 when train k-1 does not depart; is a binary variable, which is 1 if train k departs from station i at time t, and 0 otherwise; is a binary variable, which is 1 if train k-1 departs from station i at time t, and 0 otherwise.

[0069] The remaining passenger capacity is:

[0070]

[0071] wherein pc k,i is the number of passengers that train k can accommodate at station i, i.e., the train capacity; C m is the initial empty capacity of train k.

[0072] The number of passengers getting on the train is:

[0073] pb k,i = min{pw k,i , pc k,i}, i∈S, k∈K;

[0074] wherein pb k,i is the number of passengers getting on train k at station i.

[0075] The number of passengers on the train is:

[0076]

[0077] Further, the objective function in step 202 includes a first sub-objective function and a second sub-objective function.

[0078] The first sub-objective function is:

[0079]

[0080] The second sub-objective function is:

[0081]

[0082] The objective function is:

[0083] F = w1 · F1 + w2 · F2;

[0084] Wherein, w1 and w2 are the weights of the double-objective optimization method.

[0085] Further, the train diagram model in step 203 is solved to obtain the train diagram, which can be solved by Gurobi solver or CPLEX solver, supporting mixed integer programming (MIP) and being suitable for train diagram problem.

[0086] Further, the double-depth Q network algorithm is used to generate the optimal train running speed curve according to the train historical running data, train diagram and line data in step 204, as shown in Figure 3 , which specifically includes:

[0087] In step 2041, the train running reinforcement learning environment is constructed according to the train historical running data, train diagram, line data and the reward function obtained; the distance between each station, slope and train speed limit are taken as the static data of the line, and the current position, speed, acceleration and running resistance of the train are taken as the dynamic data of the train running. In the process of train running, the agent interacts with the train running reinforcement learning environment, the environment generates new train running state, reward value and state value function feedback to the agent. The agent constantly evaluates and improves the strategy through the value function, and selects the maximum action value function, feeds back the maximum action value to the train running reinforcement learning environment, and constantly updates the running condition value through such a closed loop structure, and finally selects the optimal running condition action to generate the optimal train speed curve.

[0088] In this embodiment, the state set of the train is set as:

[0089]

[0090] Wherein, x represents the current position of the train, and Xi where d represents the distance between two stations. This ratio represents the relative position of the train at the current time; v represents the speed of the train at the current time, V i min represents the minimum allowable speed of train i, and this ratio reflects the relative speed of the train at the current time relative to its minimum allowable speed, which can indicate whether the train is running at a low speed; t represents the current time, and T p represents the planned running time, which generally refers to the time required for the train to run between two adjacent stations according to the plan, and this ratio represents the current relative time, which can reflect the position of the train on the time axis; V i represents the speed limit of the train, and V max represents the maximum allowable speed of the train, and this ratio reflects the relative speed of the train at the current time relative to its maximum allowable speed, which can indicate whether the train is running at a high speed or approaching its speed limit.

[0091] The action set of the train is set as:

[0092] a = {0, 1, 2, 3};

[0093] where 0, 1, 2, 3 represent traction, cruising, coasting, and braking in turn, which are the possible actions that the agent (i.e., the on-board controller) can output.

[0094] The reward function of the train is set as:

[0095]

[0096] At each current time state of the train, the train running reinforcement learning environment gives a reward value to the action selected by the agent, which is used to evaluate the quality of the action. Among them, w3·ΔT represents the punctuality reward, which reflects the difference between the train arrival time and the target time, ΔT represents the difference between the actual arrival time of the train and the target arrival time (which can be positive or negative), when the train arrives ahead of schedule, ΔT is negative; when the train arrives late, ΔT is positive, and the weight w3 is a positive number, so when the train deviates from the target time, whether it is ahead of schedule or delayed, a negative reward (i.e., a penalty) will be obtained. represents the energy saving reward, which reflects the rate of change of energy consumption of the train during operation, represents the rate of change of energy consumption over time, which can be positive or negative (but generally energy consumption is increasing, so this item is usually negative), and the weight w4 is a positive number, so when the energy consumption increases, a negative reward (i.e., a penalty) will be obtained, if the train takes energy-saving measures to reduce energy consumption (although this may be less common in actual situations), a positive reward will be obtained. comfort, which reflects the change in train acceleration, i.e., the rate of change of acceleration, which affects the comfort of passengers, denotes the absolute value of the rate of change of acceleration, and dt is the time interval. w6·(X i -x) denotes the position accuracy reward, which reflects the position accuracy of the train in a certain section, X i denotes the target position of the train in section i, x denotes the current position of the train, and the sign of the weight w6 depends on the optimization goal: if the train is expected to be rewarded when it approaches the target position, w6 is positive; if the train is expected to be punished when it deviates from the target position, w6 is negative. Here, w6 will be set to a positive number to encourage the train to maintain accurate position in a certain section.

[0097] In step 2042, a train speed curve optimization network is constructed based on a double deep Q network.

[0098] In step 2043, the train speed curve optimization network is trained based on the train operation reinforcement learning environment to obtain a trained train speed curve optimization network.

[0099] Further, the train speed curve optimization network includes an evaluation network and a target network; the determination process of the trained train speed curve optimization network is as follows:

[0100] Initializing the evaluation network and the target network is a process of giving initial values to the weights and biases in the deep neural network or policy model. Good network parameter initialization is crucial for the convergence speed and final performance of the algorithm. Through proper initialization, the problem of gradient vanishing or explosion can be avoided, and the stability of the training process can be improved. Xavier initialization is adopted to select appropriate initial value range according to the number of network layers and the characteristics of the activation function, so as to ensure that the signal can be effectively transmitted in the forward propagation and backward propagation process.

[0101] Based on the train operation reinforcement learning environment, the current time state of the train is determined; the current time state includes the current time relative position, the current time first relative speed, the current time relative time and the current time second relative speed.

[0102] The current time state is input into the evaluation network to determine the current time action; the evaluation network calculates the future return expectation (i.e., Q value) of each action according to the current state and the action list, and then selects an action to execute according to the greedy strategy (or ε-greedy strategy), which is usually the action with the highest Q value (or has a certain probability to randomly select an action for exploration).

[0103] According to the current time action, the train operation reinforcement learning environment is used to calculate the current time reward value and determine the next time state.

[0104] The current time reward value and the next time state are input into the target network to calculate the target Q value. The evaluation network and the target network are updated according to the current time state, the current time action, the target Q value and the loss function; specifically, the evaluation network predicts a Q value according to the current state and the selected action, and the difference between the target Q value and the Q value predicted by the evaluation network is used to calculate the loss function, and the formula is as follows:

[0105] yloss=E[(r+γmaxQ(s,b';θ-)-Q(s,b;θ)) 2 ];

[0106] Wherein, r is the reward value, γ is the discount factor, b represents all actions predicted by the DQN network, b' represents all actions output by the DQN network, θ represents the parameters used for prediction of the DQN network, θ - represents the parameters of the target network, and Q(s,b;θ) is the Q value predicted by the evaluation network.

[0107] Repeat the above steps until a predetermined number of iterations is reached or the loss function converges to a predetermined threshold, and obtain the trained train speed curve optimization network.

[0108] Step 2044, based on the trained train speed curve optimization network, the optimal train running speed curve is obtained, as shown in Figure 4 .

[0109] To evaluate the performance of the virtual coupling train energy optimization method, we conducted three experiments. The first experiment compared the fixed coupling strategy and the virtual coupling strategy. The second experiment studied the influence of the weight coefficient in the objective function to understand the trade-off between energy consumption and service quality. The last experiment evaluated the impact of passenger demand changes on the performance of the virtual coupling train energy optimization method.

[0110] Different train coupling number evaluation: compare the performance of the fixed coupling strategy and the virtual coupling strategy. Specifically, a total of 20 instances are constructed, with the number of trains K increasing from 14 to 52. For the fixed coupling strategy, long coupled train formations are used; while for the virtual coupling strategy, different types of train formations are allowed. In each instance, Gurobi (version 10.0.1) is used to solve the model to achieve the optimal solution (with an error of no more than 0.5%).

[0111] Different train adjustment strategy evaluation: In the objective function of the train diagram model, the weight factors w1 and w2 represent the contribution of total delay and train operation cost respectively. To study the impact of different strategies on train coupling, the number of trains is set to 14, and the weight ratio w 1 / w2 is set to 0.05, 0.06, 0.07, 0.10, 0.14, 0.25, 0.5, 1 respectively.Figure 5 The performance of the fixed marshalling coupling and the virtual marshalling coupling strategies, including passenger waiting time and overall energy consumption of trains, is demonstrated in these 8 instances. Since trains have different preferences in performance, different train adjustment strategies need to be adopted according to limited resources and passenger demand.

[0112] Evaluation under different passenger demand: In order to quantitatively evaluate the effect of different passenger demand, a total of 20 instances are constructed, with the number of trains K increasing from 14 to 52. In each instance, we set w1 = 1 and w2 = 10 in the objective function. Figures 6 and 7 respectively demonstrate the performance of the fixed marshalling coupling and the virtual marshalling coupling strategies, including passenger waiting time and energy consumption of trains, in these 20 instances under different passenger demand (off-peak and peak, respectively).

[0113] The application further provides an application scenario of the virtual marshalling train energy-saving optimization method. Specifically, the virtual marshalling train energy-saving optimization method provided in the embodiment can be applied in a virtual marshalling train energy-saving optimization scenario. The virtual marshalling train energy-saving optimization scenario includes a train running speed curve optimization link and a train running speed curve application link; the train running speed curve optimization link is used to obtain an optimal train running speed curve according to line data and train historical running data of different types of trains; the train running speed curve application link is used to make the train travel according to the optimal train running speed curve; and the virtual marshalling train energy-saving optimization method belongs to the train running speed curve optimization link.

[0114] In an exemplary embodiment, a computer device, which can be a server or a terminal, is provided, and an internal structure diagram of the computer device can be as shown in Figure 8 The computer device includes a processor, a memory, an input / output interface (I / O), and a communication interface. The processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The database of the computer device is used to store processing data. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals through network connection. The computer program is executed by the processor to implement a virtual marshalling train energy-saving optimization method.

[0115] Those skilled in the art can understand that, Figure 8The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0116] In an exemplary embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, and the processor implementing the steps in the above method embodiments when executing the computer program.

[0117] In an exemplary embodiment, a computer readable storage medium is provided, storing a computer program, and the computer program implementing the steps in the above method embodiments when executed by a processor.

[0118] In an exemplary embodiment, a computer program product is provided, including a computer program, and the computer program implementing the steps in the above method embodiments when executed by a processor.

[0119] 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 for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant regulations.

[0120] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, database or other medium used in the embodiments provided in the present 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 storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0121] The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a blockchain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0122] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combinations of the technical features do not exist contradictory, they should be considered as the scope of the present application.

[0123] The principles and implementation modes of the present application are described by applying specific examples herein, and the above-mentioned embodiments are only used to help understand the method and its core idea of the present application; meanwhile, for those skilled in the art, according to the idea of the present application, the specific implementation mode and application range can be changed. In conclusion, the content of the present application should not be understood as a limitation.

Claims

1. A virtual consist train energy optimization method, characterized in that, The virtual marshalling train energy-saving optimization method comprises: acquiring line data and train historical operation data of different types of trains, and calculating train operation energy consumption of different types of trains according to the line data and the train historical operation data of different types of trains; the line data comprises distances between stations, slopes and train speed limits, and the train historical operation data comprises positions, speeds, accelerations and running resistances of trains at current time; based on a virtual marshalling technology and a mixed integer programming method, a train diagram model is constructed according to the train operation energy consumption; the train diagram model comprises constraint conditions and a target function, the constraint conditions comprise train timetable constraints, passenger flow constraints and train coupling constraints, and the target function is a function taking balancing passenger waiting time and the train operation energy consumption as targets; the train diagram model is solved to obtain a train diagram; an optimal train operation speed curve is generated by using a double-depth Q network algorithm according to the train historical operation data, the train diagram and the line data.

2. The virtual consist train energy optimization method of claim 1, wherein, The train operation energy consumption of different types of trains is calculated according to the line data and the train historical operation data of different types of trains, and the specific process is as follows: The train operation energy consumption of different types of trains is calculated according to the following formula: ； wherein, is the speed of the train of type at the station at time t = 0; is the speed of the train of type at the station after time to arrive at the station with mass ; is the speed of the train of type at the station at time t; is the running time of the train at the station ; is the mass of the train of type ; is the minimum speed limit at the station at time t; is the set of stations; is the set of virtual consist train types; is the set of train quantities; ； ； wherein, is the speed of the train at time t+1; is the distance of the train at time t+1; is the speed of the train at time t; is the distance of the train at time t; is the ratio of the braking force output by the train per unit time; is the ratio of the traction force output by the train per unit time; is the braking force of the train at the station at a speed of is the basic resistance of the train at the station at a speed of is the gradient force of the station at a speed of is the traction force of the train at the station at a speed of is the gradient force of the station at a speed of is the traction force of the train at the station at a speed of is the traction force of the train at the station at a speed of ； ； ； wherein, is the tractive effort of the train at a speed of ; is the braking effort of the train at a speed of ; is the maximum tractive effort; is the maximum tractive power; is the tractive break speed limit; is the maximum braking effort; is the maximum braking power; is the braking break speed limit; is the basic resistance of the train at a speed of ; , , is the Davis equation coefficient representative of the train parameter characteristics; ； wherein, is the total energy consumption of the train, composed of the cumulative traction energy consumption generated between each station; is the running distance of the train between stations; denotes the distance between two stations.

3. The virtual consist train energy optimization method of claim 1, wherein, The train timetable constraints comprise train interval operation time constraints and train arrival and departure time constraints; The train interval operation time constraints are as follows: ； wherein, is a train at a station departure time; is a train at a station departure time; is a safety headway between consecutive trains; is a train; is a set of train numbers; is a set of stations; The train arrival and departure time constraints are as follows: ； wherein for a train at a station of arrival time; for a train at a station of departure time; for a train from a station of departure time; The train coupling constraints are as follows: ； wherein, is a decision variable, ensuring that the train can only choose one type of marshalling.

4. The virtual consist train energy optimization method of claim 1, wherein, The passenger flow constraints comprise passengers getting off, waiting passenger quantity, remaining passenger capacity, passenger boarding number and passenger number on trains; The passengers getting off are as follows: ； wherein, is the number of trains, is the number of passengers, is the number of passengers, is the number of passengers, is the number of passengers, is the number of passengers; is the ratio, is the set of numbers of trains, is the set of stations. The waiting passenger quantity is as follows: ； in, For at the station Waiting for the train The number of passengers; For at the station The previous train The number of passengers who failed to board the train; For the previous train Arrive at the station after leaving The number of passengers; For the station The percentage of passengers remaining on the train at time t; As a slack variable, the train When departing from [t, t+1], from the station The offset of departure time relative to t; its value is 0 if no departure occurs. As a binary variable, if the train at time t At the station The value is 1 if the activity starts, and 0 otherwise. As a binary variable, if the train at time t At the station The value is 1 if the activity starts, and 0 otherwise. The remaining passenger capacity is as follows: ； wherein, for the train at the station the number of passengers the train can accommodate, i.e. the train capacity; for the train the initial empty capacity; The passenger boarding number is as follows: ； wherein for a train at a station the number of passengers boarding the train; The passenger number on trains is as follows: 。 5. The virtual consist train energy optimization method of claim 1, wherein, The target function comprises a first sub-target function and a second sub-target function; The first sub-target function is as follows: ； The second sub-target function is as follows: ； The target function is as follows: ； wherein, and are weights of the bi-objective optimization method; is the number of passengers waiting for a train at a station ; is the number of passengers waiting for a train at a station ; is the number of passengers boarding a train at a station ; is the departure time of a train at a station ; is the departure time of a train at a station ; is the departure time of a train at a station ; is a decision variable ensuring that a train can only choose one type of marshalling; is the total energy consumption of a train, consisting of the cumulative traction energy generated between stations; is a train; is a set of train numbers; is a set of stations.

6. The virtual consist train energy optimization method of claim 1, wherein, The optimal train operation speed curve is generated by using a double-depth Q network algorithm according to the train historical operation data, the train diagram and the line data, and the specific process comprises: a train operation reinforcement learning environment is constructed according to the train historical operation data, the train diagram, the line data and an obtained reward function; a train speed curve optimization network is constructed based on a double-depth Q network; the train speed curve optimization network is trained based on the train operation reinforcement learning environment to obtain a trained train speed curve optimization network; the optimal train operation speed curve is obtained based on the trained train speed curve optimization network.

7. The virtual consist train energy optimization method of claim 6, wherein, The reward function is as follows: ； wherein, represents punctuality reward; represents energy saving reward; represents comfort reward, dt is time interval; represents position accuracy reward; , , and is a weight of the reward function; represents a difference between actual arrival time of the train and target arrival time; represents energy consumption; represents speed of the train at current time; represents current time; represents distance between two stations; represents position of the train at current time.

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 virtual marshalling train energy-saving optimization method in any one of claims 1-7.

9. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the virtual marshalling train energy-saving optimization method in any one of claims 1-7.

Images