Power system recovery method and device, storage medium and electronic equipment

Abstract

The invention discloses a power system recovery method and device, a storage medium and electronic equipment. The method comprises the steps that the current system state of a power system is obtained, the power system is in a fault state, and new energy output exists in the power system; determining a first control action based on the current system state; after the first control action is executed, the updated system state of the power system is obtained, the action value of the first control action is determined, the updated system state comprises the updated new energy output state, and the action value is used for representing the recovery effect of the execution action; determining and executing a second control action based on the updated system state and the action value of the first control action; and adopting a mode of determining and executing a second control action, and repeatedly carrying out recovery control until the power system is converted from the fault state to the normal state. The technical problem that the accuracy of power supply recovery is not ideal in the prior art is solved.

Description

Technical Field

[0001] The present application relates to the field of power systems, and in particular to a power system recovery method, device, storage medium and electronic equipment. Background Art

[0002] As the penetration rate of renewable energy continues to increase, the complexity and uncertainty of power grid operation are increasing, posing huge challenges to the safe operation of the power grid. At the same time, a large number of power electronic equipment have been introduced into the distribution network, greatly improving the operational flexibility of the power grid.

[0003] When constructing an optimization model for a power supply restoration solution in related technologies, it is necessary to know grid information such as the faulty node, the network structure after fault protection action, the location of the smart flexible switch (Soft Open Point, SOP), and the optimization is based on scenarios or random optimization. In actual operation, it is difficult to cope with the volatility and intermittency of renewable resources, resulting in inaccurate control of grid fault recovery.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present application provide a power system recovery method, device, storage medium and electronic device to at least solve the technical problem of unsatisfactory accuracy of power supply recovery in the related art.

[0006] According to one aspect of an embodiment of the present application, there is provided a method for restoring an electric power system, comprising: obtaining a current system state of the electric power system, wherein the electric power system is in a fault state and there is a new energy output in the electric power system; determining a first control action based on the current system state, wherein the first control action is used to control an adjustable device in the electric power system; after executing the first control action, obtaining an updated system state of the electric power system, and determining an action value of the first control action, wherein the updated system state includes an updated state of the new energy output, and the action value is used to represent the restorative effect of executing the action; determining and executing a second control action based on the updated system state and the action value of the first control action; and repeating recovery control by determining and executing the second control action until the electric power system changes from a fault state to a normal state.

[0007] Optionally, determining and executing a second control action based on the updated system state and the action value of the first control action includes: updating a deep reinforcement learning algorithm based on the action value of the first control action to obtain an updated deep reinforcement learning algorithm; inputting the updated system state into the updated deep reinforcement learning algorithm for processing to determine the action values ​​corresponding to multiple candidate actions; determining the second control action with the highest action value among the multiple candidate actions, and executing the second control action.

[0008] Optionally, determining the action value of the first control action includes: using a reward function to determine the action value of the first control action, wherein the reward function is determined based on an objective function and satisfaction of constraints, the objective function is used to represent the load recovery amount of the power system, and the constraints are used to constrain the system state of the power system during the execution of the recovery action.

[0009] Optionally, the power system includes multiple nodes, and the multiple nodes respectively correspond to different weight coefficients, and the weight coefficients represent the severity of power supply interruption of the corresponding nodes. The method also includes: determining the node recovery amount corresponding to the multiple nodes based on the power demands and weight coefficients respectively corresponding to the multiple nodes; determining the load recovery amount based on the node recovery amount corresponding to the multiple nodes.

[0010] Optionally, the constraints include at least: constraints on node voltage in the power system, constraints on line current, capacity constraints on smart flexible switches SOP, constraints on connection on / off states, constraints on upper limits on output of distributed power sources DG, and no loop constraints in the power system.

[0011] Optionally, obtaining the current system state of the power system includes: obtaining a node voltage, a branch current, a current active load, a current reactive load, a faulty active load in a fault state, and a faulty reactive load in the power system; and determining the current system state based on the node voltage, the branch current, the current active load, the current reactive load, the faulty active load, and the faulty reactive load.

[0012] Optionally, the adjustable device includes an intelligent flexible switch SOP and a controlled distributed power source DG, and the first control action is determined based on the current system state, including: based on the current system state, determining the on-off state of the SOP, controlling the first power setting on both sides of the SOP, and controlling the second power setting of the DG, as the first control action.

[0013] According to another aspect of an embodiment of the present application, there is provided an electric power system recovery device, comprising: a first state acquisition module, for acquiring a current system state of the electric power system, wherein the electric power system is in a fault state and there is a new energy output in the electric power system; a first action determination module, for determining a first control action based on the current system state, wherein the first control action is used to control an adjustable device in the electric power system; a second state acquisition module, for acquiring an updated system state of the electric power system after executing the first control action, and determining an action value of the first control action, wherein the updated system state includes an updated state of the new energy output, and the action value is used to represent the recovery effect of executing the action; a second action determination module, for determining and executing a second control action based on the updated system state and the action value of the first control action; and a cyclic recovery module, for repeatedly performing recovery control by determining and executing the second control action until the electric power system is transformed from a fault state to a normal state.

[0014] According to another aspect of an embodiment of the present application, a non-volatile storage medium is provided, wherein the non-volatile storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded by a processor and executing any one of the power system recovery methods described above.

[0015] According to another aspect of an embodiment of the present application, there is provided an electronic device, comprising: one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement any one of the power system recovery methods described.

[0016] In an embodiment of the present application, by obtaining the current system state of the power system, wherein the power system is in a fault state and there is a new energy output in the power system; based on the current system state, determining a first control action, wherein the first control action is used to control the adjustable equipment in the power system; after executing the first control action, obtaining the updated system state of the power system, and determining the action value of the first control action, wherein the updated system state includes the updated state of the new energy output, and the action value is used to indicate the recovery effect of the execution action; based on the updated system state and the action value of the first control action, determining and executing a second control action; using the method of determining and executing the second control action, repeatedly performing recovery control until the power system changes from a fault state to a normal state. The purpose of more effectively utilizing the new energy output for power recovery is achieved, and the technical effect of improving the accuracy and efficiency of power supply recovery is achieved, thereby solving the technical problem of unsatisfactory accuracy of power supply recovery in the related art. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0018] Figure 1 is a flowchart of an optional power system restoration method provided according to an embodiment of the present application;

[0019] Figure 2 is a schematic diagram of an algorithm processing of an optional power system restoration method provided according to an embodiment of the present application;

[0020] Figure 3 is a schematic diagram of action selection of an optional power system restoration method provided according to an embodiment of the present application;

[0021] Figure 4 is a schematic flow chart of an optional power system restoration method provided according to an embodiment of the present application;

[0022] Figure 5 It is a schematic diagram of an optional power system recovery device provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0023] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0024] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0025] For the convenience of description, some nouns or terms involved in the embodiments of the present application are explained below:

[0026] The intelligent flexible switch SOP is a new type of intelligent power electronic device installed at the traditional tie switch. It can change the distribution network from the traditional closed-loop design open-loop operation to a flexible closed-loop operation. It can perform real-time and precise power control at the feeder level to support the intelligent operation of the system and flexibly switch between multiple operating states. It has a faster response speed and stronger power flow regulation ability, which can effectively balance the load between feeders and improve power supply capacity.

[0027] Distributed Generation (DG) refers to small modular, environmentally compatible independent power sources with power ranging from several kilowatts to 50MW (megawatts). These power sources are usually close to the user side and can meet the special needs of end users. DG energy forms are diverse, and can be a variety of energy forms based on new energy, such as photovoltaic power generation, wind power generation, hydropower generation, fuel cells, electric vehicle batteries, etc.; it can also be diesel power generation; it can also exist in the form of combined heat, cooling and electricity, providing heat energy to users on site, improving the utilization efficiency and flexibility of DG.

[0028] Related technologies generally take the maximum load of power restoration as the goal, use the switching of tie switches and DG power as control variables, and use network operation safety as a constraint to build a power restoration model. For uncertainty processing, scenario-based methods or random optimization methods are generally used. The utilization rate of energy for power restoration is irrational, and there are inaccurate problems in power restoration control.

[0029] In response to the above-mentioned problems, an embodiment of the present application provides an optional method embodiment for power system recovery. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0030] Figure 1 is a flow chart of an optional power system restoration method according to an embodiment of the present application, such as Figure 1 As shown, the method comprises the following steps:

[0031] Step S102, obtaining the current system state of the power system, wherein the power system is in a fault state and there is new energy output in the power system;

[0032] It is understandable that in order to restore the power system in a faulty state, the real-time data of the power system is collected and processed, which may include the topology of the power grid, the voltage and current conditions of each node, the load level, and the output status of new energy (such as wind power and photovoltaic power), etc. The above data together constitute the current system status of the power system, which can provide a basis for data processing for subsequent decision-making.

[0033] In an optional embodiment, obtaining the current system state of the power system includes: obtaining the node voltage, branch current, current active load, current reactive load, faulty active load in a fault state, and faulty reactive load in the power system; determining the current system state based on the node voltage, branch current, current active load, current reactive load, faulty active load, and faulty reactive load.

[0034] It can be understood that the current system state of the power system can be obtained based on the node voltage, branch current, current active load, current reactive load, faulty active load in fault state, and faulty reactive load. The node voltage reflects the potential difference of each node in the power system, the branch current indicates the current size on each branch in the power network, the current active load and the current reactive load respectively indicate the active power and reactive power currently consumed in the power system, which are the basis for evaluating the system load level and performing load management. Faulty active load and faulty reactive load are loads that are temporarily unable to supply power due to system failures (such as equipment failures, line failures, etc.). Combined with the load level and the output of new energy, it provides a data basis for what kind of recovery measures the power system should take in the current system state.

[0035] Optionally, at time t, the power system state is composed of three types of parameters that are mainly related to the dynamic critical load restoration (DCLR) decision, including node voltage Vi, branch current I ij , active and reactive load P load , Q load , and the loss of active and reactive load P shed , Q shed , which has the characteristics of "continuous state space and discrete action space". Since the state space S contains continuous variables of power type, S has the characteristic of "continuity". The state of the power system at time t is expressed as:

[0036]

[0037] Among them, s t represents the current system state of the power system at time t, represents the current active load of node i in the power system at time t, represents the current reactive load of node i at time t, represents the fault active load of node i in the power system at time t, represents the fault reactive load of node i at time t, V i,t represents the node voltage of node i at time t, I ij,t represents the branch current between node i and node j at time t, Γ represents the time set, N b Represents a collection of nodes in a power system.

[0038] Step S104, determining a first control action based on the current system state, wherein the first control action is used to control an adjustable device in the power system;

[0039] It can be understood that after obtaining the current system status, analysis and processing are performed based on the current system status to determine the first control action, which aims to initially improve the system status or alleviate the impact of the fault by adjusting the adjustable equipment in the power system (such as SOP, tie switches, DG, etc.).

[0040] In an optional embodiment, the adjustable device includes an intelligent flexible switch SOP and a controlled distributed power source DG, and determines a first control action based on a current system state, including: determining the on-off state of the controlled SOP based on the current system state, controlling a first power setting on both sides of the SOP, and controlling a second power setting of the DG as the first control action.

[0041] It is understood that adjustable devices may include smart flexible switches (SOPs) and distributed power sources (DGs) to achieve dynamic adjustment based on the current system state. SOPs can quickly switch their on and off states according to system needs. SOPs are also provided with converters on both sides, which can control the setting of power transmission to ensure the reasonable distribution of power in the power grid. DG power output control can adjust the power output of DG according to the real-time needs of the power grid and the available output of DG, and can maximize the use of renewable energy to achieve clean and efficient use of energy. Through the optimization of power settings on both sides of SOP and the precise control of DG power output, the resources in the power grid can be configured and utilized more reasonably and efficiently, reducing the waste and loss of resources caused by power outages.

[0042] Optionally, the controllable variables of SOP include the active power transmitted by each end converter and the reactive power output. Whether it is the inter-station SOP or the inter-feeder SOP, the PQ (active power and reactive power) control strategy is adopted, which can be expressed as follows:

[0043]

[0044] Among them, ij are the identifiers of different nodes in the power system, represents the active power (inflow or outflow) transmitted by node i through SOP, represents the active power (outflow or inflow) transmitted by node j through SOP, represents the active power (inflow or outflow) lost by node i during the SOP process, The active power lost (outflow or inflow) at node j during the SOP process, represents the loss coefficient of node i, represents the loss coefficient of node j, represents the reactive power (inflow or outflow) transmitted by node i through SOP, represents the reactive power (outflow or inflow) transmitted by node j through SOP, represents the apparent power transmitted by node i through SOP, represents the apparent power transmitted by node j through SOP.

[0045] Step S106, after executing the first control action, obtaining an updated system state of the power system, and determining an action value of the first control action, wherein the updated system state includes the updated state of the new energy output, and the action value is used to indicate the recovery effect of executing the action;

[0046] It can be understood that after executing the first control action, the real-time data of the power system is collected again to obtain the updated system status, that is, the updated system status. The updated system status reflects the impact of the control action on the system, including the update of the new energy output. At the same time, the action value of executing the first control action is evaluated, that is, the positive role played by executing the action on the recovery of the power system. By executing control actions and observing their effects, the power system can adjust subsequent recovery strategies in real time. Adjusting control actions according to real-time changes in real-time system status improves the adaptability of the system, especially for power systems with new energy output. Since the new energy output is irregular and will change in real time, executing recovery decisions in steps will help improve the adaptability to new energy scenarios.

[0047] Step S108, determining and executing a second control action based on the updated system state and the action value of the first control action;

[0048] It can be understood that the second control action is determined based on the updated system state and the action value of the first control action. The second control action further improves the recovery state of the power system after the previous step and accelerates the recovery process. By continuously iterating control actions, the system can gradually approach the optimal recovery strategy.

[0049] In an optional embodiment, based on the updated system state and the action value of the first control action, a second control action is determined and executed, including: based on the action value of the first control action, updating the deep reinforcement learning algorithm to obtain an updated deep reinforcement learning algorithm; inputting the updated system state into the updated deep reinforcement learning algorithm for processing to determine the action values ​​corresponding to multiple candidate actions; determining a second control action with the highest action value among multiple candidate actions, and executing the second control action.

[0050] It can be understood that the state of the power system will be updated with time, external conditions (such as load changes, fluctuations in renewable energy output, etc.), and the execution of previous control actions. The effect of the first control action is evaluated, that is, its action value is calculated, which reflects the degree of improvement of the power system state by executing the first control action, or can be considered as the contribution to the realization of the restoration goal. Based on the action value feedback of the first control action, the deep reinforcement learning algorithm is updated. Through continuous trial and error and adjustment, the deep reinforcement learning algorithm can gradually learn the optimal control strategy to be adopted under different system states. The updated deep reinforcement learning algorithm contains more experience and knowledge and can better adapt to the changes in the power system state. The updated system state is input into the updated deep reinforcement learning algorithm, and multiple candidate control actions are generated according to the updated system state. The action value that may be brought after the execution of each candidate action is predicted, which provides an important basis for decision-making. Among the multiple candidate actions, the one with the highest action value is selected as the second control action, and the second control action is executed, which ensures that the control action taken is optimal (or locally optimal) under the updated system state, which is conducive to maximizing the realization of the restoration goal.

[0051] The real-time update of the deep reinforcement learning algorithm and the value evaluation of candidate actions facilitate the rapid and accurate determination of the optimal control action, thereby improving the efficiency and accuracy of the recovery control decision. Through continuous experience and knowledge, the system can adapt to the changing state of the power system, so that the system can maintain high stability and reliability in the face of complex and changing power environments (such as changes in the output of new energy).

[0052] Optionally, a deep Q-network (DQN) is used as a model-free deep reinforcement learning algorithm that does not require additional prior knowledge in the process of searching for the optimal strategy. The Q-network uses a multi-layer perceptron model, and the Q-network parameters are iteratively adjusted during the training process to gradually reduce the gap between the action value function and the target value function. The input of the Q-network is the system state, and the output is the action value of each action. The action corresponding to the highest action value in the output is the optimal load recovery strategy. By adopting the highest value continuous control action, the power flow transfer optimization can be achieved.

[0053] Figure 2 is a schematic diagram of an algorithm processing of an optional power system restoration method provided in an embodiment of the present application, such as Figure 2 As shown, the Q network can be expressed as Q(s, a; θ)≈Q*(s, a), where Q* is the optimal target action value, Q is the action value, and θ is the weight parameter, which means that Q is close to Q*. At each iteration, the distance between the actual action value and the optimal action value can be expressed by the loss function F:

[0054] F(θ μ )=E((Q * (s',a',θ')-Q(s,a,θ)) 2 )

[0055] For the loss function F(θ), the parameter θ can be iteratively updated based on optimization algorithms such as stochastic gradient descent.

[0056] Optionally, after the fault section is removed, the power flow optimization problem of the remaining system belongs to the dynamic critical load restoration problem (DCLR), which has the characteristics of "continuous state space-discrete action space" and can be modeled as a Markov decision process (MDP). The recovery process can be divided into a series of single-stage problems to be solved one by one. The decision of each stage considers the impact of a certain number of future decisions, and the solutions of each stage constitute the optimal decision sequence of the whole process.

[0057] Figure 3 is a schematic diagram of action selection of an optional power system recovery method provided according to an embodiment of the present application, such as Figure 3 As shown in the figure, the Markov decision model MDP consists of a five-tuple {S, A, P, r, γ}, where S is the state space; A is the action space, P(s'|s, a) is the probability of reaching state s' from state s after taking action a, r(s, a, s') is the immediate reward of reaching state s' from state s after taking action a, and γ is the discount factor, 0≤γ≤1. t, t+1, t+2, etc. are the marks for performing different actions. For the DCLR mathematical model of the dynamic critical load recovery problem, the Markov decision MDP model of the dynamic critical load recovery problem DCLR is constructed by defining elements such as intelligent agents, interactive environments, control actions, system states, and reward functions.

[0058] For the agent and the interactive environment, the agent in the DCLR problem will assume the role of DSO and formulate a load recovery strategy from the global perspective of the distribution system. In the DCLR (Distribution System Contingency Load Restoration) problem, the agent will assume the role of DSO (Distribution System Operator) and formulate a load recovery strategy from the global perspective of the distribution system. DSO is the entity responsible for operating, managing and maintaining the distribution system (that is, the part of the power network responsible for delivering electricity from the transmission system to the end user). The environment that interacts with the agent is a distribution system simulator with functions such as action execution, state analysis and reward feedback.

[0059] The control action a taken by the agent at time t t It is expressed as:

[0060]

[0061] Among them, w,t is a binary state variable, indicating the operation of switch w at time t, o w,t =0 means closing operation, o w,t =1 indicates disconnection operation, O indicates the set of state variables, N w represents the set of switch nodes, and Γ represents the set of moments. Since the action space A consists of finite binary state variables, A has the characteristic of "discreteness". They represent the active power (P) and reactive power (Q) output by the kth distributed generation DG at time t, respectively. DG Represents the collection of distributed generation DG. They are respectively the active power and reactive power output by the converters on both sides of the h-th SOP (identified by 1 and 2) at time t.

[0062] At time t, the system state consists of three types of parameters that are mainly related to the dynamic critical load recovery problem DCLR decision, including the ith node voltage Vi, the branch current Iij between the ith node and the jth node, and the active and reactive load P load , Q load , and the loss of active and reactive load P shed , Q shed Since the state space S contains continuous variables of the power type, S has the characteristic of "continuous", and the system state is expressed as:

[0063]

[0064] Among them, V i,trepresents the voltage of the i-th node at the t-th moment, I ij,t represents the branch current between the ith node and the jth node at the tth moment, represents the active load of the ith node at the tth moment, represents the reactive load of the ith node at the tth moment, represents the active power lost by the ith node at the tth moment, represents the reactive power lost by the ith node at the tth moment, N b Represents a collection of branches.

[0065] In an optional embodiment, determining the action value of the first control action includes: using a reward function to determine the action value of the first control action, wherein the reward function is determined based on an objective function and satisfaction of constraints, the objective function is used to represent the load recovery amount of the power system, and the constraints are used to constrain the system state of the power system during the execution of the recovery action.

[0066] It can be understood that the objective function is used to measure the contribution of the first control action to the load recovery of the power system, and is designed to be a function that can quantify the load recovery amount. For example, the higher the value of the objective function, the greater the contribution of the control action to the system load recovery. Constraints are a set of rules or restrictions that must be met to ensure that the constraints are a set of rules or restrictions that must be met, ensuring that the power system maintains a safe, stable and economical operating state during the recovery process. The reward function determines the reward value of the execution action based on the satisfaction of the objective function and the constraints. If a control action can significantly improve the load recovery amount and comply with all constraints, it will receive a higher reward value. Conversely, if the control action has limited contribution to the load recovery amount or violates certain constraints, its reward value will be reduced accordingly. Through the above processing, a basis is provided for decision-making in the power system recovery process, which can help the intelligent agent evaluate the value of different control actions from a global perspective and formulate the optimal load recovery strategy.

[0067] In an optional embodiment, the constraints include at least: constraints on node voltage in the power system, constraints on line current, capacity constraints on the smart flexible switch SOP, constraints on the connection on / off state, constraints on the upper limit of the output of the distributed power source DG, and no-loop constraints in the power system.

[0068] It can be understood that the above constraints are a set of rules or restrictions that must be met to ensure that the power system maintains a safe and stable operating state during the recovery process. The node voltage constraint, each node in the power system (such as generators, transformers, loads and other connection points) has its allowable range of voltage. The node voltage constraint ensures that the voltage level is neither too high nor too low to avoid equipment damage or reduced operating efficiency, which can be determined by the rated voltage of the equipment and the allowable voltage deviation.

[0069] Line current constraint refers to the current flowing through the power system transmission line. Since the line has thermal capacity limitations and electrical insulation limitations, the current must be kept within a safe range. Excessive current may cause line overheating, insulation damage, or even fire. Therefore, constraining line current is an important measure to protect the safety of the power grid.

[0070] The capacity constraint of the smart flexible switch SOP can achieve flexible power flow control in the power system. However, the capacity of the SOP is limited, that is, the power that can be transmitted is limited. Therefore, the capacity constraint of the SOP needs to be considered during planning and operation to ensure that it does not exceed the design capacity, thereby protecting the equipment and maintaining the stability of the system.

[0071] The tie-line (or tie-switch) in the power system is used to connect different grid areas or subsystems. The on-off state of the tie-line directly affects the structure and power flow distribution of the grid to ensure the stability and reliability of the grid under a specific operating mode.

[0072] Distributed power generation DG output limit constraints, with the rapid development of renewable energy, distributed power generation in the power system accounted for an increasingly higher proportion. However, the output of distributed power generation is affected by many factors (such as weather, equipment conditions, etc.), and its capacity is limited. In the power system, it is necessary to constrain the output limit of distributed power generation to prevent its excessive power generation from causing impact on the power grid.

[0073] There is no loop constraint in the power system, which means that the topology of the power grid avoids the formation of loops, which will lead to complex power flow distribution, difficult protection configuration and power grid stability problems, so as to maintain the structural clarity and stability of the power grid.

[0074] Optionally, constraints may also be included, such as: power balance constraints, which are used to ensure that the restored system can maintain a balance between supply and demand and avoid overload or underload. Equipment capacity constraints, which are used to consider the capacity limitations of equipment such as generators, transformers, and lines to prevent equipment from overloading. Network topology constraints, which are used to ensure the rationality of the recovery path, such as considering switch status, line connectivity, etc. Safety and stability constraints, which are used to ensure that system parameters such as voltage and frequency are within a safe and stable range during the recovery process. Operation rule constraints, which are used to consider specific operation rules or regulations in the operation of the power system, such as recovery order, priority, etc.

[0075] Optionally, the reward function directly affects the decision-making behavior of the agent. Therefore, the design of the reward function needs to comprehensively consider the impact of the recovery goal (ie, the objective function) and the constraints on the decision.

[0076] r t =(c1α 1,t -C2α 2,t -C3α 3,t -C4α 4,t -C5α 5,t -C6α 6,t -C7α 7,t )

[0077] Among them, c1α 1,t represents the reward for restoring the load at time t, -c2α 2,t represents the penalty for violating the node voltage constraint at time t, -c3α 3,t represents the penalty for violating the line current constraint at time t, -c4α 4,t represents the penalty for violating the SOP capacity constraint, -c5α 5,t represents the penalty for violating the tie switch state constraint, -c6α 6,t Indicates the penalty for violating the DG output upper limit constraint, -c7α 7,t Represents the penalty for violating the radial constraint of the system.

[0078] In an optional embodiment, the power system includes multiple nodes, and the multiple nodes respectively correspond to different weight coefficients, and the weight coefficients represent the severity of power supply interruption of the corresponding nodes. The method also includes: determining the node recovery amount corresponding to the multiple nodes based on the power demands and weight coefficients respectively corresponding to the multiple nodes; determining the load recovery amount based on the node recovery amount corresponding to the multiple nodes.

[0079] It can be understood that the nodes in the power system are assigned different weight coefficients to quantify the severity that each node may suffer when the power supply is interrupted, so that when restoring power or allocating resources, those nodes that have a greater impact or are more critical to the system can be given priority. Based on the power demand of each node and its corresponding weight coefficient, the node recovery amount of each node can be calculated, and the corresponding recovery operation can be performed according to the formulated recovery strategy. This may include starting backup generators, adjusting the grid structure, repairing damaged equipment, etc., to ensure that the power system can be restored to normal as soon as possible. Through this method of determining the recovery amount based on node weight coefficients and power demand, the power system can perform recovery work more efficiently and orderly when the power supply is interrupted, thereby minimizing losses and impacts.

[0080] Optionally, there are many ways to establish the objective function. For example, according to different requirements for power supply reliability and the impact of power outages on personal safety, supply and demand, and economy, different loads can generally be divided into three levels: primary load, secondary load, and tertiary load. Among them, the interruption of primary and secondary loads will lead to serious consequences. Therefore, these important loads have priority restoration, so different priority loads are assigned different weight coefficients, and ultimately the distribution network power supply restoration load is maximized, which is solved in the form of objective function f, that is:

[0081]

[0082] Among them, the maximum value of f is the maximum restored load of the distribution network in the entire cycle, j is the jth node of the distribution system, is the active power demand of the load at node j at time t, ω j is the weight coefficient of the jth node load, 0<ω j <1, N is the total number of nodes in the power distribution system, and T is the time period.

[0083] The constraints can include: power flow constraints, node voltage and branch current constraints, branch transmission capacity linearization constraints, and radial topology constraints. The optional mathematical expressions are as follows:

[0084] For power flow constraints, the active power and reactive power balance equations at node i at time t can be expressed as:

[0085]

[0086] in, represents the injected active power of node i at time t, k:i→k represents the sum of all active powers from node i to load (power source) k, P ik,t,s represents the active power from node i to load (power source) k, i:i→j represents the sum of the active powers of all branches flowing from node i to node j, P ij,t,srepresents the power between branches ij, I ij,t,s represents the current between branches ij, r ij and x ij Respectively represent the resistance and reactance of the branch between the ij nodes, g j represents the real part of the admittance at node i, V j,t,s represents the voltage of node j at time t.

[0087] Ohm's law on branch ij at time T can be expressed as:

[0088] (V j,t,s ) 2 =(V i,t,s ) 2 -2(P ij,t,s r ij +Q ij,t,s x ij )+(I ij,t,s ) 2 ((r ij ) 2 +(x ij ) 2 )

[0089] The current of each branch can be expressed as:

[0090]

[0091] Among them, the total active power and reactive power injection equations at node i at time t can be expressed as:

[0092]

[0093] in, are the active power and reactive power of DG at node i at time t, respectively. represents the active load at node i at time t, represents the reactive load at node i at time t.

[0094] Node voltage and branch current constraints:

[0095]

[0096] in, represents the minimum value of the node voltage at node i, represents the maximum value of the node voltage at node i,

[0097] V i,t represents the node voltage at node i at time t, Represents the maximum current of branch ij.

[0098] SOP capacity constraints:

[0099]

[0100] For the selection of the transfer contact switch, there are only two states for each contact switch: 0 and 1.

[0101]

[0102] For radial topology constraints, which are used to constrain the power grid topology to not form a loop, it can be expressed as:

[0103]

[0104] Among them, i→j∈C k Represents any two nodes i and j, b ij,t represents the path from node i to j, C and P are the set of all rings in the network and the set of all paths between root nodes, respectively. k , P k It represents any path between any ring and the root node in the power grid. E, N, and R all represent sets, which respectively represent the set consisting of all possible branches, the set consisting of all nodes, and the set consisting of all root nodes.

[0105] Step S110, repeatedly performing recovery control by determining and executing a second control action until the power system changes from a fault state to a normal state.

[0106] It can be understood that the control system continuously repeats the process of steps S106 to S108, that is, continuously executes control actions, evaluates effects, and determines new actions until the power system completely changes from a fault state to a normal state. By continuously iterating and optimizing the control strategy, it is ensured that the power system can stably and reliably resume normal operation after a fault.

[0107] Through the above step S102, the current system state of the power system is obtained, wherein the power system is in a fault state and there is new energy output in the power system; step S104, based on the current system state, a first control action is determined, wherein the first control action is used to control the adjustable equipment in the power system; step S106, after executing the first control action, the updated system state of the power system is obtained, and the action value of the first control action is determined, wherein the updated system state includes the state of the updated new energy output, and the action value is used to indicate the recovery effect of the execution action; step S108, based on the updated system state and the action value of the first control action, a second control action is determined and executed; step S110, recovery control is repeated by determining and executing the second control action until the power system changes from a fault state to a normal state. The purpose of more effectively utilizing new energy output for power recovery is achieved, and the technical effect of improving the accuracy and efficiency of power supply recovery is achieved, thereby solving the technical problem of unsatisfactory accuracy of power supply recovery in the related art.

[0108] According to the above optional embodiment, the present application also provides an optional implementation mode: Figure 4 is a schematic flow chart of an optional power system restoration method provided according to an embodiment of the present application, such as Figure 4 As shown, in step S1, the distribution network topology, renewable energy output, load curve, and system fault status are obtained to build a power system simulation model.

[0109] Step S2, initialize the experience replay pool, action value network Q and target action value network Q.

[0110] Step S3: Based on the current system state, construct the system action space and select action a based on the ε-greedy strategy. t .

[0111] Step S4, executing action a in the power system simulation model t , environmental feedback reward r t and the next state s t+1 , the experience (s t , a t , r t ,s t+1 ) is stored in the experience replay pool.

[0112] Step S5, according to the loss function, use the stochastic gradient descent method to update the network Q parameters; after every C updates, replace the network Q with the network Q.

[0113] Step S6: if the iteration time is up or the optimal action value is reached, go to step S7; otherwise, t=t+1, go to step S3.

[0114] Step S7, save the network Q and end the training. Use the trained model to perform restoration decision processing on the power system.

[0115] Through deep learning algorithms, system data is collected, trained, and decisions are made to improve the system's ability to cope with uncertain factors and enhance the power supply recovery capability during system failures.

[0116] In this embodiment, a power system recovery device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the terms "module" and "device" can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0117] According to an embodiment of the present application, there is also provided an apparatus embodiment for implementing a power system restoration method. Figure 5 is a schematic diagram of a power system restoration device according to an embodiment of the present application, such as Figure 5 As shown, the above-mentioned power system recovery device includes: a first state acquisition module 502, a first action determination module 504, a second state acquisition module 506, a second action determination module 508, and a cycle recovery module 510. The device is described below.

[0118] The first state acquisition module 502 is used to acquire the current system state of the power system, wherein the power system is in a fault state and there is new energy output in the power system;

[0119] A first action determination module 504, connected to the first state acquisition module 502, for determining a first control action based on the current system state, wherein the first control action is used to control an adjustable device in the power system;

[0120] The second state acquisition module 506 is connected to the first action determination module 504 and is used to obtain an updated system state of the power system after executing the first control action, and to determine an action value of the first control action, wherein the updated system state includes the updated state of the new energy output, and the action value is used to indicate the recovery effect of the execution action;

[0121] A second action determination module 508, connected to the second state acquisition module 506, for determining and executing a second control action based on the updated system state and the action value of the first control action;

[0122] The cyclic recovery module 510 is connected to the second action determination module 508 and is used to repeatedly perform recovery control by determining and executing the second control action until the power system changes from a fault state to a normal state.

[0123] In a power system recovery device provided by an embodiment of the present application, a first state acquisition module 502 is set to obtain the current system state of the power system, wherein the power system is in a fault state and there is new energy output in the power system; a first action determination module 504 is connected to the first state acquisition module 502 and is used to determine a first control action based on the current system state, wherein the first control action is used to control an adjustable device in the power system; a second state acquisition module 506 is connected to the first action determination module 504 and is used to obtain an updated system state of the power system after executing the first control action, and to determine an action value of the first control action, wherein the updated system state includes the updated state of the new energy output, and the action value is used to indicate the recovery effect of the execution action; a second action determination module 508 is connected to the second state acquisition module 506 and is used to determine and execute a second control action based on the updated system state and the action value of the first control action; a cyclic recovery module 510 is connected to the second action determination module 508 and is used to repeatedly perform recovery control by determining and executing the second control action until the power system changes from a fault state to a normal state. The purpose of more effectively utilizing the output of new energy for power restoration is achieved, the technical effect of improving the accuracy and efficiency of power restoration is realized, and the technical problem of unsatisfactory accuracy of power restoration existing in related technologies is solved.

[0124] It should be noted that the above modules can be implemented by software or hardware. For example, for the latter, it can be implemented in the following ways: the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0125] It should be noted that the first state acquisition module 502, the first action determination module 504, the second state acquisition module 506, the second action determination module 508, and the loop recovery module 510 correspond to steps S102 to S110 in the embodiment, and the examples and application scenarios implemented by the above modules and the corresponding steps are the same, but are not limited to the contents disclosed in the above embodiments. It should be noted that the above modules can be run in a computer terminal as part of the device.

[0126] It should be noted that the optional or preferred implementation of this embodiment can refer to the relevant description in the embodiment, which will not be repeated here.

[0127] The above-mentioned power system recovery device may also include a processor and a memory. The first state acquisition module 502, the first action determination module 504, the second state acquisition module 506, the second action determination module 508, the cycle recovery module 510, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize the corresponding functions.

[0128] The processor includes a kernel, which retrieves the corresponding program unit from the memory. There can be one or more kernels. The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one storage chip.

[0129] An embodiment of the present application provides a non-volatile storage medium having a program stored thereon, which implements a power system recovery method when executed by a processor.

[0130] The embodiment of the present application provides an electronic device, which includes a processor, a memory, and a program stored in the memory and can be run on the processor. When the processor executes the program, the following steps are implemented: obtaining the current system state of the power system, wherein the power system is in a fault state and there is a new energy output in the power system; based on the current system state, determining a first control action, wherein the first control action is used to control an adjustable device in the power system; after executing the first control action, obtaining an updated system state of the power system, and determining an action value of the first control action, wherein the updated system state includes the updated state of the new energy output, and the action value is used to indicate the recovery effect of the execution action; based on the updated system state and the action value of the first control action, determining and executing a second control action; using the method of determining and executing the second control action, repeatedly performing recovery control until the power system changes from a fault state to a normal state. The device in this article can be a server, a PC, etc.

[0131] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program that is initialized with the following method steps: obtaining a current system state of the power system, wherein the power system is in a fault state and there is new energy output in the power system; determining a first control action based on the current system state, wherein the first control action is used to control an adjustable device in the power system; after executing the first control action, obtaining an updated system state of the power system, and determining an action value of the first control action, wherein the updated system state includes the state of the updated new energy output, and the action value is used to represent the recovery effect of the execution of the action; determining and executing a second control action based on the updated system state and the action value of the first control action; and repeating recovery control by determining and executing the second control action until the power system changes from a fault state to a normal state.

[0132] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0133] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0134] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0135] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0136] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0137] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0138] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0139] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0140] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0141] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for restoring a power system, characterized in that: include: Acquire a current system state of a power system, wherein the power system is in a fault state and there is new energy output in the power system; Based on the current system state, determining a first control action, wherein the first control action is used to control an adjustable device in the power system; After executing the first control action, obtaining an updated system state of the power system and determining an action value of the first control action, wherein the updated system state includes an updated state of the new energy output, and the action value is used to represent a recovery effect of executing the action; determining and executing a second control action based on the updated system state and the action value of the first control action; The recovery control is repeated in a manner of determining and executing the second control action until the power system changes from a fault state to a normal state.

2. The method according to claim 1, characterized in that The determining and executing a second control action based on the updated system state and the action value of the first control action comprises: Based on the action value of the first control action, the deep reinforcement learning algorithm is updated to obtain an updated deep reinforcement learning algorithm; Inputting the updated system state into the updated deep reinforcement learning algorithm for processing to determine the action values ​​corresponding to the multiple candidate actions; The second control action having the highest action value is determined among the multiple candidate actions, and the second control action is executed.

3. The method according to claim 1, characterized in that The determining the action value of the first control action includes: A reward function is used to determine the action value of the first control action, wherein the reward function is determined based on an objective function and the satisfaction of constraint conditions, the objective function is used to represent the load recovery amount of the power system, and the constraint conditions are used to constrain the system state of the power system during the execution of the recovery action.

4. The method according to claim 3, characterized in that The power system includes a plurality of nodes, each of which corresponds to a different weight coefficient, and the weight coefficient indicates the severity of power supply interruption of the corresponding node. The method further includes: Determine node recovery amounts corresponding to the plurality of nodes respectively based on power requirements and weight coefficients corresponding to the plurality of nodes respectively; The load recovery amount is determined based on the node recovery amounts respectively corresponding to the multiple nodes.

5. The method according to claim 3, characterized in that: The constraints include at least: constraints on node voltage in the power system, constraints on line current, capacity constraints on smart flexible switches SOP, constraints on the connection on / off state, constraints on the upper limit of the output of distributed power sources DG, and no loop constraints in the power system.

6. The method according to claim 1, characterized in that The obtaining of the current system state of the power system comprises: Acquiring node voltages, branch currents, current active loads, current reactive loads, faulty active loads in a faulty state, and faulty reactive loads in the power system; The current system state is determined based on the node voltage, the branch current, the current active load, the current reactive load, the fault active load, and the fault reactive load.

7. The method according to any one of claims 1 to 6, characterized in that: The adjustable device includes a smart flexible switch SOP and a control distributed power source DG, and the first control action is determined based on the current system state, including: Based on the current system state, determining to control the on / off state of the SOP, controlling the first power setting of both sides of the SOP, and controlling the second power setting of the DG is performed as the first control action.

8. A power system restoration device, characterized in that: include: A first state acquisition module is used to acquire the current system state of the power system, wherein the power system is in a fault state and there is new energy output in the power system; A first action determination module, configured to determine a first control action based on the current system state, wherein the first control action is used to control an adjustable device in the power system; a second state acquisition module, configured to acquire an updated system state of the power system after executing the first control action, and determine an action value of the first control action, wherein the updated system state includes an updated state of the new energy output, and the action value is used to represent a recovery effect of executing the action; A second action determination module, configured to determine and execute a second control action based on the updated system state and the action value of the first control action; A cyclic recovery module is used to repeatedly perform recovery control in a manner of determining and executing the second control action until the power system changes from a fault state to a normal state.

9. A non-volatile storage medium, characterized in that: The non-volatile storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the power system restoration method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: include: One or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the power system recovery method described in any one of claims 1 to 7.

Images