Urban rail traction power supply system power supply capacity improvement method and system

Abstract

The application provides a power supply capacity improvement method and system for a city rail traction power supply system, belongs to the technical field of city rail traction power supply systems, obtains traction substation incoming line power, train speed, train power and train position information data of the city rail traction power supply system, pre-processes the collected data, establishes a city rail traction power supply system simulation model, predicts traction substation peak power in a fixed time period from the present to the future, performs traction substation peak power prediction parameter regulation, generates control instructions with the most parameters, and regulates and controls the converter device. The application reduces the traction substation incoming line peak power, effectively improves the power supply capacity, coordinates and controls the working parameters of the above controllable converter device, dynamically balances the power distribution among multiple devices under the premise of ensuring the safe operation of the system, and finally achieves the goal of improving the overall power supply capacity of the system.

Description

Technical Field

[0001] This invention relates to the field of urban rail transit traction power supply system technology, specifically to a method and system for improving the power supply capacity of urban rail traction power supply systems. Background Technology

[0002] In urban rail transit traction power supply systems, the limited capacity of external power sources under a distributed power supply architecture, combined with the pulsed load characteristics caused by frequent train starts and stops and traction braking, constitutes the core contradiction in system operation. During peak operating periods, the instantaneous power generated by the dense departure of trains easily approaches or exceeds the capacity of external power sources and the rated load capacity of switching equipment. This can lead to safety hazards such as equipment overheating and protection malfunctions, and directly limits the expansion space for the number of trains operating on the line. The current power supply system adopts a combined architecture of diode rectifier units and controllable converters (including energy feeders, bidirectional converters, and energy storage devices). Although it has basic power supply capabilities, the fixed droop coefficient control strategy has inherent limitations: on the one hand, it lacks multi-device collaborative decision-making capabilities, making it difficult to achieve dynamic optimization of power distribution ratios, resulting in power imbalance among multiple devices, with some converters overloaded while other devices have idle capacity; on the other hand, the traditional static control mode that relies on local voltage / current feedback is difficult to actively smooth out the peak power of the external power source and cannot adapt to the rapid load fluctuations caused by sudden changes in train operating conditions. To this end, the present invention optimizes the controllable converter's regulation strategy to achieve active adjustment of the incoming peak power and dynamic power balance distribution, thereby improving the system's power supply capacity while maintaining the existing external power supply capacity. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for improving the power supply capacity of urban rail traction power supply systems, so as to solve at least one of the technical problems existing in the background art.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] In a first aspect, the present invention provides a method for improving the power supply capacity of an urban rail traction power supply system, comprising:

[0006] Acquire data on the incoming power of the traction substation, train speed, train power, and train position of the urban rail traction power supply system;

[0007] The collected data was filled with missing values ​​and outliers were removed.

[0008] Establish a simulation model of the urban rail traction power supply system, including the converter units in the traction substation, the step-down substation, the DC traction network, and the train;

[0009] Predict the peak power of the traction station from the present to a fixed time period in the future based on a pre-trained prediction model;

[0010] Parameter regulation for peak power prediction of traction substations is based on a pre-trained regulation model.

[0011] The optimal parameters are used to generate control commands to regulate the converter.

[0012] As a further limitation of the first aspect of the present invention, the peak power prediction parameter regulation of the traction substation based on the pre-trained regulation model includes: initially determining a fixed simulation period t, first predicting the peak power of the traction substation in the time period [0, t], and determining whether it exceeds the limit. If it exceeds the limit and occurs at time t1, then using reinforcement learning to regulate the converter unit parameters of the traction substation at that time. After the regulation is completed, since this regulation will affect the previously predicted state of each traction substation, it is necessary to re-predict for [t1, t], and repeat the prediction and regulation process until the peak power of the traction substation at all times [0, t] is controlled within a reasonable range, and the regulation command for this period is output.

[0013] As a further limitation of the first aspect of the present invention, the mechanism-driven simulation model of the urban rail traction power supply system includes an equivalent model of the actual topology and a basic model simulation algorithm; the equivalent topology model includes the connection information between the traction substation and the step-down substation and the traction network; the basic model simulation algorithm includes train traction calculation and DC power flow calculation algorithms, the simulation step size is set to 1 second, and the simulation period is a fixed time period that can be arbitrarily set.

[0014] As a further definition of the first aspect of the present invention, the state is defined as the current set of key parameters of the system; the continuous action space is defined as the adjustment amount of relevant parameters of the traction substation converter unit; the reward is designed as a composite reward function according to different objectives, wherein the penalty is based on the number of traction substations exceeding the limit, and the more substations there are, the greater the penalty.

[0015] As a further limitation of the first aspect of the present invention, when the peak power is less than before the regulation, a positive reward is given; otherwise, a negative reward is given. When an over-limit traction station is added after regulation, a penalty is given. When the inter-station distribution balance is high after regulation, a reward is given. When the voltage and current change suddenly after regulation, a penalty is given, but the weight is relatively low. Since the dimensions of each sub-reward function are different, they need to be normalized first and then added together to obtain the total reward function.

[0016] As a further limitation of the first aspect of the present invention, the strategy adopts an Actor-Critic architecture: the Actor network outputs the adjustment amount of the action parameters of the traction substation's strain flow unit combination; the Critic network is a global value assessment network that integrates the status and action information of all traction substations to estimate the value.

[0017] Secondly, the present invention provides a power supply capacity enhancement system for urban rail traction power supply systems, comprising:

[0018] The acquisition module is used to acquire data on the incoming power of the traction substation, train speed, train power, and train position information of the urban rail traction power supply system.

[0019] The preprocessing module is used to fill in missing values ​​and remove outliers from the collected data;

[0020] A module is established to create a simulation model of the urban rail traction power supply system, including the converter units in the traction substation, the step-down substation, the DC traction network, and the train.

[0021] The prediction module is used to predict the peak power of the traction station from the present to a fixed time period in the future based on a pre-trained prediction model.

[0022] The control module is used to control parameters based on the framework of "prediction-control-re-prediction-re-control".

[0023] The control module is used to generate control commands from the optimal parameters to regulate the converter.

[0024] Thirdly, the present invention provides a non-transitory computer-readable storage medium for storing computer instructions, which, when executed by a processor, implement the method for improving the power supply capacity of the urban rail traction power supply system as described in the first aspect.

[0025] Fourthly, the present invention provides a computer device including a memory and a processor, wherein the processor and the memory communicate with each other, the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the method for improving the power supply capacity of the urban rail traction power supply system as described in the first aspect.

[0026] Fifthly, the present invention provides an electronic device, comprising: a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to cause the electronic device to execute instructions for implementing the method for improving the power supply capacity of the urban rail traction power supply system as described in the first aspect.

[0027] The beneficial effects of this invention are as follows: By collecting system operation data in real time, a basic equivalent model of the traction power supply system is established and a simulation algorithm is developed. Combined with reinforcement learning technology, the no-load voltage and droop coefficient of the controllable converter are dynamically optimized, thereby reducing the peak power of the traction substation's incoming line and effectively improving power supply capacity. The power supply system adopted is a distributed architecture, with each traction substation powered by an independent external power source. Each substation is equipped with a composite power supply unit containing one diode rectifier unit and one controllable converter (including a bidirectional converter, energy storage device, or energy feeder device). By coordinating and controlling the operating parameters of the aforementioned controllable converter, the power distribution among multiple devices is dynamically balanced while ensuring safe system operation, ultimately achieving the goal of improving the overall power supply capacity of the system.

[0028] The advantages of additional aspects of the invention will be set forth more clearly in the following description or will be learned by practice of the invention. Attached Figure Description

[0029] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 This is a flowchart of the method for improving the power supply capacity of the urban rail traction power supply system according to an embodiment of the present invention.

[0031] Figure 2 This is a schematic diagram of the equivalent circuit of the DC traction network according to an embodiment of the present invention.

[0032] Figure 3 This is an equivalent circuit diagram of the DC traction power supply system described in an embodiment of the present invention.

[0033] Figure 4 This is a flowchart of the model-based traction station power prediction process described in an embodiment of the present invention.

[0034] Figure 5 This is a flowchart of the reinforcement learning model training method according to an embodiment of the present invention.

[0035] Figure 6 This is a schematic diagram illustrating the execution principle of the method for improving the power supply capacity of the urban rail traction power supply system according to an embodiment of the present invention. Detailed Implementation

[0036] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0037] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0038] It should also be understood that terms such as those defined in general dictionaries should be understood to have meanings consistent with their meanings in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as here.

[0039] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, and / or groups thereof.

[0040] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0041] To facilitate understanding of the present invention, the present invention will be further explained and described below with reference to the accompanying drawings and specific embodiments. However, the specific embodiments do not constitute a limitation on the embodiments of the present invention.

[0042] Those skilled in the art should understand that the accompanying drawings are merely schematic diagrams of embodiments, and the components in the drawings are not necessarily essential for implementing the present invention.

[0043] Example 1

[0044] In this embodiment 1, a power supply capacity enhancement system for urban rail traction power supply system is first provided, including: an acquisition module for acquiring data on the incoming power of the traction substation, train speed, train power, and train position information of the urban rail traction power supply system; a preprocessing module for completing missing values ​​and removing outliers from the collected data; a modeling module for establishing a simulation model of the urban rail traction power supply system, including the converter units, step-down substations, DC traction network, and trains in the traction substation; a prediction module for predicting the peak power of the traction substation from the present to a fixed time period in the future based on a pre-trained prediction model; a regulation module for parameter regulation based on a "prediction-regulation-re-prediction-re-regulation" framework; and a control module for generating control commands from the optimal parameters to regulate the converter device.

[0045] In this embodiment, the above-described system is used to implement a method for improving the power supply capacity of an urban rail traction power supply system, including: acquiring data on the incoming power of the traction substation, train speed, train power, and train position information; completing missing values ​​and removing outliers from the collected data; establishing a simulation model of the urban rail traction power supply system, including the converter units, step-down substations, DC traction network, and trains in the traction substation; predicting the peak power of the traction substation from the present to a fixed time period in the future based on a pre-trained prediction model; adjusting the peak power prediction parameters of the traction substation based on a pre-trained control model; and generating control commands from the optimal parameters to regulate the converter device.

[0046] The method of adjusting the peak power prediction parameters of the traction substation based on the pre-trained control model includes: initially determining a fixed simulation period t, first predicting the peak power of the traction substation in the time period [0, t] and determining whether it exceeds the limit. If it exceeds the limit and occurs at time t1, then reinforcement learning is used to adjust the converter unit parameters of the traction substation at that time. After the adjustment is completed, since this adjustment will affect the previously predicted state of each traction substation, it is necessary to re-predict for [t1, t] and repeat the prediction and adjustment process until the peak power of the traction substation in all times [0, t] is controlled within a reasonable range, and the control command for this period is output.

[0047] The simulation model of the urban rail traction power supply system based on mechanism includes an equivalent model of the actual topology and a basic model simulation algorithm. The equivalent topology model includes the connection information between the traction substation, the step-down substation, and the traction network. The basic model simulation algorithm includes train traction calculation and DC power flow calculation algorithms. The simulation step size is set to 1 second, and the simulation period is a fixed time period that can be set arbitrarily.

[0048] The state is defined as the current set of key parameters of the system; the continuous action space is defined as the adjustment amount of relevant parameters of the traction substation converter unit; the reward is designed as a composite reward function based on different objectives, where penalties are imposed based on the number of traction substations exceeding the limit, with more substations resulting in greater penalties. A positive reward is given when the peak power is less than before regulation, otherwise a negative reward is given; a penalty is imposed when a new traction substation exceeds the limit after regulation; a reward is given when the inter-station distribution balance is high after regulation; a penalty is imposed when voltage and current change abruptly after regulation, but with a relatively low weight; since the dimensions of each sub-reward function are different, they need to be normalized before being added together to obtain the total reward function.

[0049] The strategy adopts an Actor-Critic architecture: the Actor network outputs the adjustment of the action parameters of the traction substation's strain flow unit combination; the Critic network is a global value assessment network that integrates the status and action information of all traction substations to estimate the value.

[0050] Example 2

[0051] This embodiment proposes a method for improving the power supply capacity of a distributed urban rail traction power supply system. This method involves real-time acquisition of system operation data, establishing a basic equivalent model of the traction power supply system, developing a simulation algorithm, and combining reinforcement learning technology to dynamically optimize the no-load voltage and droop coefficient of the controllable converter, thereby reducing the peak power of the traction substation's incoming lines and effectively improving the power supply capacity. The targeted power supply system adopts a distributed architecture, with each traction substation powered by an independent external power source. Each substation is equipped with a composite power supply unit containing one diode rectifier unit and one controllable converter (including a bidirectional converter, energy storage device, or energy feeder device). By coordinating and controlling the operating parameters of the aforementioned controllable converter, the power distribution among multiple devices is dynamically balanced while ensuring the safe operation of the system, ultimately achieving the goal of improving the overall power supply capacity of the system. Specific implementation steps are as follows: Figure 1 As shown.

[0052] The system collects real-time information on the train's current speed, power, and location, and obtains fixed parameters of the traction power supply system and preset train operation information. The fixed parameters of the traction power supply system include converter unit parameters, traction network unit equivalent resistance, station location, track information, and operation organization information.

[0053] The collected data is processed, including missing value completion, outlier removal, and noise reduction of the raw data.

[0054] The modeling objects of the urban rail traction power supply system include traction substations, step-down substations, DC traction networks, and trains. The equivalent models of each part are established as follows.

[0055] The traction substation mainly focuses on the equivalent state of the converter units within the substation. The converter units include rectifier units and controllable converter devices.

[0056] The rectifier unit is equivalent to an ideal voltage source in series with an equivalent resistor, and its parameters cannot be adjusted.

[0057] U rect =U0-I·R rect (1)

[0058] The traction state of the controllable converter is represented as follows, and all parameters are controllable.

[0059] U bus =U0-I·k droop (2)

[0060] The step-down function is equivalent to a PQ node, and its specific mathematical expression is as follows.

[0061]

[0062] In the formula, P ds Q ds These represent active power and reactive power, respectively, where η represents the load factor, and S... ds Indicates transformer capacity. This represents the power factor.

[0063] The traction network is equivalent to a resistance model that varies with length. The train, depending on its operating state, is equivalent to a power source or power load. The specific power value is calculated based on the known train operation plan according to the train traction regulations. For example... Figure 2 Chinese R n1 R n2 This represents the equivalent resistance on both sides of the train, i.e., the equivalent resistance between the rails and the overhead contact line; P represents the train power, the specific value of which is obtained from traction calculations. The final DC traction power supply system and the train's equivalent model are as follows: Figure 3 As shown. Figure 3 In the diagram, U0 represents the traction substation voltage, and R... d R represents the equivalent resistance within the traction depot. The two resistors, connected in series, together represent the equivalent model of the traction depot. i (i = 1, 2, 3, ...) represent the equivalent resistance of the traction network, P i (i = 1, 2, 3, ...) represent the power of all trains in the current DC power supply system.

[0064] The collected and processed train data is used to correct the corresponding parameters in the original train operation plan and input into the simulation model established in step three. Through train traction calculations and traction network DC power flow calculations, the peak power information of each traction substation within a fixed future time period is obtained. Subsequent parameter adjustments are then based on these predicted values. The detailed flowchart is shown below. Figure 4As shown, the collected train operation data, preset train operation plan, and track conditions are first input into the simulation model established in the previous step. The track conditions include the gradient and curve information of the train operation track required for train traction calculations. Substituting this information into the simulation model, two basic calculations—train traction calculation and power flow calculation—can output the peak traction power information for a future period.

[0065] This study focuses on the controllable converter of a traction substation whose peak power exceeds its rated power, regulating two parameters: its no-load voltage and droop coefficient. A framework of "predicting [0,t] – regulating (time t1) – re-predicting [t1,t] – re-regulating (time t2)" is designed. Specifically, a fixed simulation period t is initially determined. The peak power of the traction substation during the time period [0,t] is predicted, and it is determined whether it exceeds the limit. If it exceeds the limit and occurs at time t1, reinforcement learning is used to regulate the converter parameters of the traction substation at that time. After regulation, since this regulation affects the previously predicted states of each traction substation, it is necessary to re-predict [t1,t], repeating the prediction and regulation process until the peak power of the traction substation at all times [0,t] is controlled within a reasonable range. The regulation command for this time period is then output. Figure 5 The diagram illustrates the specific steps of the model training process. Step 1: Initialize the environment and agent, and set the simulation period. Step 2: Increase the time step. Step 3: Predict the peak power of the traction station within a fixed time period based on the model. Step 4: Determine if any traction station exceeds the limit; if so, proceed to Step 5; otherwise, proceed to the final step to end training. Step 5: Obtain the current state, i.e., the current set of key system parameters, as shown in Equation 4. Step 6: The policy network outputs actions based on the current state, as shown in Equation 5. Step 7: The system executes the actions from Step 6. Step 8: After executing the actions, the environment model generates feedback information. Step 9: Calculate the reward based on the generated feedback information; the total reward calculation is shown in Equation 12. Step 10: Update the model based on the reward to ensure the model is closer to reality. Step 11: Determine if the current time has reached the set maximum value; if so, training ends; otherwise, proceed to Step 2.

[0066] Parameter tuning is implemented using model-based reinforcement learning. The following describes the design of the environment, actions, states, and rewards in the tuning process using the reinforcement learning method.

[0067] Environmental Model: A mechanism-driven simulation model of the urban rail traction power supply system, including an equivalent model of the actual topology and basic model simulation algorithms. The equivalent topology model includes the connection information between the traction substation, the step-down substation, and the traction network. Figure 3The equivalent circuit diagram is shown. The basic model simulation algorithm includes train traction calculation and DC power flow calculation algorithms. The simulation step size is set to 1 second, and the simulation period is a fixed time period, which can be set arbitrarily.

[0068] State: The state is defined as the current set of key parameters of the system, including the following variables as shown in Table 1.

[0069]

[0070] Table 1 Explanation of State Variables

[0071]

[0072] Action: The continuous action space is defined as the adjustment amount of relevant parameters of the traction substation converter unit.

[0073] ΔU0: No-load voltage adjustment of the controllable converter (range: ±10% of rated voltage)

[0074] Δk droop Adjustment amount of droop coefficient for controllable converter (range: ±0.1)

[0075]

[0076] Rewards: Rewards are designed using composite reward functions based on different objectives.

[0077] ① Main penalty item

[0078] Penalties are imposed based on the number of oversized traction units; the more units there are, the heavier the penalty.

[0079] R1 = -10 * Sub_num (6)

[0080] In the formula, Sub_num represents the number of oversized traction stations.

[0081] ② Incoming peak power reduction reward: When the peak power is less than before the control, a positive reward is given; otherwise, a negative reward is given.

[0082]

[0083] In the formula, P peak_j new P represents the peak power of the j-th traction substation's incoming line after adjustment. peak_j old This represents the peak power of the j-th traction substation's incoming line before the control is adjusted.

[0084] ③ Side effect punishment

[0085] If any new traction stations exceeding the limits are added after regulation, penalties will be imposed.

[0086] R3 = -30 * Sub_num_new (8)

[0087] In the formula, Sub_num_nuw represents the number of newly added over-limit traction stations after adjustment.

[0088] ④ Power Balance Reward

[0089] A high degree of balance in the distribution among stations after regulation will be rewarded.

[0090]

[0091] In the formula, Inter-station distribution balance.

[0092] ⑤ System stability reward

[0093] Sudden changes in voltage and current after regulation are penalized, but the weight is relatively low.

[0094] R5=-0.1*(ΔU+ΔI) (10)

[0095] Since the sub-reward functions have different dimensions, they need to be normalized before being added together to obtain the total reward function, which is expressed as follows:

[0096]

[0097] In the formula, R i 'Represents the normalized function value of each sub-reward function.

[0098] Strategy: Adopt the Actor-Critic architecture.

[0099] Actor Network: Input state, output adjustment of action parameters of the traction unit combination. This network uses the Proximal Policy Optimization (PPO) algorithm, the specific expression of which is as follows. The PPO Clip mechanism can ensure smooth parameter adjustment.

[0100] L CLIP (θ)=E[min(r(θ)A(s,a),clip(r(θ),1-ε,1+ε)A(s,a))] (12)

[0101] In the formula, r(θ) represents the policy ratio, which is used to measure the magnitude of policy updates. If the value is greater than 1, it means that the new policy is more inclined to choose the action, and vice versa. A(s,a) represents the evaluation of the value of the current action. This advantage function quantifies the value of adjusting a certain traction parameter compared to the average policy. ε is the pruning threshold, which is used to limit the maximum magnitude of policy updates and is set to 0.1 to 0.3.

[0102] Critic Network: A global value assessment network that integrates the status and action information of all traction stations to estimate value. This network uses the TD3 (Twin Delayed DDPG) algorithm, as shown in the following expression.

[0103]

[0104] In the formula, r represents the immediate reward; Υ represents the discount factor, set to 0.95 to make the strategy focus more on long-term stability; Q i π represents the Q-value estimate of the target Critic network for the next state s'. The double Q network takes the minimum value to avoid overestimation; π is the action generated by the target Actor network. Noise ε is added to enhance the exploration; ε represents the action noise, which is generally limited to the range of [-0.2, 0.2].

[0105] like Figure 6 As shown, the central controller updates the droop coefficient and no-load voltage based on the reinforcement learning algorithm, and generates control commands for the controllable converter, which are then sent to the converter in each traction substation to realize the control of the controllable converter.

[0106] Example 3

[0107] This embodiment 3 provides a non-transitory computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, they implement the method for improving the power supply capacity of the urban rail traction power supply system as described above. The method includes:

[0108] Acquire data on the incoming power of the traction substation, train speed, train power, and train position of the urban rail traction power supply system;

[0109] The collected data was filled with missing values ​​and outliers were removed.

[0110] Establish a simulation model of the urban rail traction power supply system, including the converter units in the traction substation, the step-down substation, the DC traction network, and the train;

[0111] Predict the peak power of the traction station from the present to a fixed time period in the future based on a pre-trained prediction model;

[0112] Parameter regulation for peak power prediction of traction substations is based on a pre-trained regulation model.

[0113] The optimal parameters are used to generate control commands to regulate the converter.

[0114] Example 4

[0115] This embodiment 4 provides a computer device, including a memory and a processor. The processor and the memory communicate with each other. The memory stores program instructions that can be executed by the processor. The processor calls the program instructions to execute the method for improving the power supply capacity of the urban rail traction power supply system as described above. The method includes:

[0116] Acquire data on the incoming power of the traction substation, train speed, train power, and train position of the urban rail traction power supply system;

[0117] The collected data was filled with missing values ​​and outliers were removed.

[0118] Establish a simulation model of the urban rail traction power supply system, including the converter units in the traction substation, the step-down substation, the DC traction network, and the train;

[0119] Predict the peak power of the traction station from the present to a fixed time period in the future based on a pre-trained prediction model;

[0120] Parameter regulation for peak power prediction of traction substations is based on a pre-trained regulation model.

[0121] The optimal parameters are used to generate control commands to regulate the converter.

[0122] Example 5

[0123] This embodiment 5 provides an electronic device, including: a processor, a memory, and a computer program; wherein, the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes instructions to implement the method for improving the power supply capacity of the urban rail traction power supply system as described above, the method including:

[0124] Acquire data on the incoming power of the traction substation, train speed, train power, and train position of the urban rail traction power supply system;

[0125] The collected data was filled with missing values ​​and outliers were removed.

[0126] Establish a simulation model of the urban rail traction power supply system, including the converter units in the traction substation, the step-down substation, the DC traction network, and the train;

[0127] Predict the peak power of the traction station from the present to a fixed time period in the future based on a pre-trained prediction model;

[0128] Parameter regulation for peak power prediction of traction substations is based on a pre-trained regulation model.

[0129] The optimal parameters are used to generate control commands to regulate the converter.

[0130] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0131] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0132] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0133] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment, whereby a series of operational steps are performed to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0134] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that, based on the technical solutions disclosed in the present invention, various modifications or variations that can be made by those skilled in the art without creative effort should be included within the scope of protection of the present invention.

Claims

1. A method for improving the power supply capacity of an urban rail traction power supply system, characterized in that, include: Acquire data on the incoming power of the traction substation, train speed, train power, and train position of the urban rail traction power supply system; The collected data was filled with missing values ​​and outliers were removed. A simulation model of the urban rail traction power supply system is established, including the converter units, step-down substations, DC traction network, and trains in the traction substations. A mechanism-driven simulation model of the urban rail traction power supply system is also established, including an equivalent model of the actual topology and basic model simulation algorithms. The equivalent topology model includes connection information between the traction substations, step-down substations, and the traction network. The basic model simulation algorithms include train traction calculation and DC power flow calculation algorithms. The simulation step size is set to 1 second, and the simulation period is a fixed time period that can be arbitrarily set. The prediction model is based on a pre-trained model to predict the peak power of the traction substation from the present to a fixed time period in the future; where the state is defined as the current set of key parameters of the system; the continuous action space is defined as the adjustment amount of relevant parameters of the converter unit of the traction substation; the reward is designed as a composite reward function according to different objectives, where the penalty is based on the number of traction substations exceeding the limit, the more the number, the greater the penalty; Parameter regulation is carried out based on the framework of "prediction-regulation-re-prediction-re-regulation"; The optimal parameters are used to generate control commands to regulate the converter.

2. The method for improving the power supply capacity of a rail traction power supply system according to claim 1, characterized in that, The peak power prediction parameter regulation of traction substations is based on a pre-trained regulation model. The process includes: initially determining a fixed simulation period t, first predicting the peak power of the traction substations in the time period [0, t] and determining whether it exceeds the limit. If it exceeds the limit and occurs at time t1, then using reinforcement learning to regulate the converter unit parameters of the traction substations at that time. After the regulation ends, since this regulation will affect the previously predicted states of each traction substation, it is necessary to re-predict for [t1, t] and repeat the prediction and regulation process until the peak power of the traction substations in all times [0, t] is controlled within a reasonable range, and the regulation command for this period is output.

3. The method for improving the power supply capacity of an urban rail traction power supply system according to claim 1, characterized in that, A positive reward is given when the peak power is lower than before the control; otherwise, a negative reward is given. A penalty is given when an over-limit traction station is added after the control. A reward will be given for achieving a high degree of balanced distribution among stations after regulation. After regulation, sudden changes in voltage and current are penalized, but with relatively low weights. Since the dimensions of each sub-reward function are different, they need to be normalized before being added together to obtain the total reward function.

4. The method for improving the power supply capacity of an urban rail traction power supply system according to claim 1, characterized in that, The strategy adopts an Actor-Critic architecture: the Actor network outputs the adjustment of the action parameters of the traction substation's strain flow unit combination; the Critic network is a global value assessment network that integrates the status and action information of all traction substations to estimate the value.

5. A power supply capacity enhancement system for urban rail traction power supply system, characterized in that, include: The acquisition module is used to acquire data on the incoming power of the traction substation, train speed, train power, and train position information of the urban rail traction power supply system. The preprocessing module is used to fill in missing values ​​and remove outliers from the collected data; The module is used to build a simulation model of the urban rail traction power supply system, including the converter units, step-down substations, DC traction network, and trains in the traction substations; the mechanism-driven simulation model of the urban rail traction power supply system includes an equivalent model of the actual topology and basic model simulation algorithms; the equivalent topology model includes the connection information between the traction substations, step-down substations, and traction network; the basic model simulation algorithms include train traction calculation and DC power flow calculation algorithms, with the simulation step size set to 1 second and the simulation period being a fixed time period that can be arbitrarily set; The prediction module is used to predict the peak power of the traction substation from the present to a fixed time period in the future based on a pre-trained prediction model. Here, the state is defined as the current set of key parameters of the system; the continuous action space is defined as the adjustment amount of relevant parameters of the converter unit of the traction substation; the reward is designed with a composite reward function according to different objectives, where the penalty is based on the number of traction substations exceeding the limit, and the more the number, the greater the penalty. The control module is used to control parameters based on the framework of "prediction-control-re-prediction-re-control"; The control module is used to generate control commands from the optimal parameters to regulate the converter.

6. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium is used to store computer instructions, which, when executed by a processor, implement the method for improving the power supply capacity of the urban rail traction power supply system as described in any one of claims 1-4.

7. A computer device, characterized in that, The system includes a memory and a processor, which communicate with each other. The memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the method for improving the power supply capacity of the urban rail traction power supply system as described in any one of claims 1-4.

8. An electronic device, characterized in that, include: The device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory to cause the electronic device to execute instructions for implementing the method for improving the power supply capacity of the urban rail traction power supply system as described in any one of claims 1-4.

Images