Active power distribution network voltage regulation and control method based on multiple time scales

By adopting a multi-time scale voltage regulation method in the active distribution network, combining load scheduling and reactive equipment, a long and short time scale voltage control strategy is formulated, and the problems of voltage fluctuations and strategy coordination are solved, and the stability of the distribution network is achieved. Operation and economic regulation.

CN119010053AActive Publication Date: 2024-11-22STATE GRID ZHEJIANG ELECTRIC POWER CO LTD JINHUA POWER SUPPLY CO +4

Patent Information

Application Number
CN202411456503.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-18
Publication Date
2024-11-22
Estimated Expiration
2044-10-18

AI Technical Summary

Technical Problem

When the existing active distribution network deals with the problem of voltage overlimiting, it cannot effectively deal with immediate short-term voltage fluctuations, and there are coordination and synchronization problems for voltage regulation methods of long-term and short-term scales, resulting in interference between equipment and control command conflicts. , affects the stable operation of the network.

Method used

Using a voltage regulation method based on multi-time scale, combining load scheduling and reactive equipment to formulate a centralized voltage control strategy for long-term time scales, and formulating a real-time voltage control strategy for short-time scales during the execution process, and real-time voltage regulation is performed through real-time voltage regulation equipment. , and periodically optimize centralized voltage control strategies to achieve coordinated and synchronized long and short time scale strategies.

Benefits of technology

Effectively solve the problem of instant and short-term voltage fluctuations in the active distribution network, avoid interference and conflicts between control strategies, ensure the stable operation of the distribution network, and ensure the real-time voltage regulation through optimized feedback regulation cycle and economical.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119010053A_ABST
    Figure CN119010053A_ABST
Patent Text Reader

Abstract

The invention provides an active power distribution network voltage regulation and control method based on multiple time scales, and the method specifically comprises the steps: obtaining load information and basic equipment information, constructing a centralized voltage control strategy model, and obtaining a centralized voltage control strategy; executing the centralized voltage control strategy, and collecting operation state information of the active power distribution network in real time; constructing a real-time voltage control strategy model, obtaining a real-time voltage control strategy, and executing the real-time voltage control strategy; and setting a feedback regulation period, obtaining a control result of the corresponding real-time voltage control strategy when the current feedback regulation period is ended, determining voltage regulation deviation information of the active power distribution network in the corresponding feedback regulation period, matching an optimization direction of the centralized voltage control strategy, and optimizing the centralized voltage control strategy. According to the invention, voltage regulation is carried out by combining the long-time scale voltage control strategy and the short-time scale voltage control strategy, coordination and synchronization of the two strategies are realized through a feedback regulation mode, and stable operation of the active power distribution network is guaranteed.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network operation control, and in particular to an active distribution network voltage control method based on multiple time scales. Background Art

[0002] The output of distributed photovoltaics is random and intermittent. As the proportion of distributed photovoltaic access continues to increase, the active distribution network with large-scale distributed photovoltaic access gradually experiences problems such as bidirectional power flow, voltage over-limit and uneven power distribution, making it difficult to ensure operational safety.

[0003] The existing voltage regulation method of the main distribution network mainly combines the reactive compensation of reactive regulation equipment such as on-load tap-changing transformers and capacitor banks, as well as the load regulation of energy storage systems and demand-side response to achieve voltage regulation for the active distribution network to solve the voltage over-limit problem.

[0004] However, this voltage regulation method belongs to long-time scale voltage regulation. Due to the limitation of its setting basis and control equipment, it is difficult to deal with the immediate short-time voltage fluctuations in the active distribution network. Although there are short-time scale voltage regulation methods that can deal with short-time voltage fluctuations in the active distribution network, the control objects used in such short-time scale voltage regulation and the control objects used in long-time scale voltage regulation have significant differences in regulation speed and characteristics. If long-time scale voltage regulation and short-time scale voltage regulation are performed at the same time, it is likely that due to the coordination and synchronization problems between the two control objects, there will be interference between devices, control command conflicts, etc., which will affect the stable operation of the active distribution network. Summary of the invention

[0005] The purpose of the present invention is to overcome the disadvantage that the prior art cannot handle the immediate short-term voltage fluctuations occurring in the active distribution network when using a long-term voltage regulation method to handle the voltage over-limit problem, and provide a voltage regulation method for an active distribution network based on multiple time scales. On the basis of formulating a long-term centralized voltage control strategy in combination with load scheduling and reactive equipment, a short-term real-time voltage control strategy is formulated. In the process of executing the centralized voltage control strategy, the real-time voltage regulation device adjusts the immediate short-term voltage fluctuations according to the real-time voltage control strategy, and periodically optimizes the centralized control strategy according to the control result of the real-time voltage control strategy, so as to realize the coordinated synchronization of the long-term centralized voltage control strategy and the short-term real-time voltage control strategy, avoid the conflict between the control results of the two control strategies, and ensure the stable operation of the active distribution network.

[0006] The purpose of the present invention is achieved through the following technical solutions: The active distribution network voltage control method based on multiple time scales includes: Step S1, obtaining load information and basic equipment information of the active distribution network; Step S2, constructing a centralized voltage control strategy model based on load information and basic equipment information, and obtaining a centralized voltage control strategy according to the centralized voltage control strategy model; Step S3, executing a centralized voltage control strategy and collecting operating status information of the active distribution network in real time; Step S4, determining the real-time voltage regulation device according to the basic device information, building a real-time voltage control strategy model based on the real-time voltage regulation device and the operation status information, obtaining the real-time voltage control strategy according to the real-time voltage control strategy model, and executing the real-time voltage control strategy; Step S5, obtaining load change information and system stability indicators based on the operating status information, setting a feedback adjustment cycle based on the load change information and the system stability indicators, and obtaining a control result of a corresponding real-time voltage control strategy at the end of the current feedback adjustment cycle; Step S6, determining the voltage regulation deviation information of the active distribution network in the corresponding feedback regulation period according to the control result of the real-time voltage control strategy, and matching the optimization direction of the centralized voltage control strategy based on the voltage regulation deviation information; Step S7, optimizing the centralized voltage control strategy according to the optimization direction of the centralized voltage control strategy obtained by matching and the voltage regulation deviation information of the active distribution network in the corresponding feedback regulation cycle, and executing the optimized centralized voltage control strategy at the beginning of the next feedback regulation cycle until the centralized voltage control strategy is completed.

[0007] Considering that long-time scale voltage regulation cannot handle the immediate short-time voltage fluctuations in the active distribution network, while executing the control strategy of long-time scale voltage regulation, the control strategy of short-time scale voltage regulation is formulated and executed. Even if there are immediate short-time voltage fluctuations, they can be solved in time to ensure the active distribution network. And the centralized control strategy can be optimized periodically according to the control results of the real-time voltage control strategy, and the coordinated synchronization of the long-time scale centralized voltage control strategy and the short-time scale real-time voltage control strategy can be achieved, avoiding interference between the control objects of the two voltage control strategies or control command conflicts, and ensuring the stable operation of the active distribution network.

[0008] Furthermore, in step S2, a centralized voltage control strategy model is constructed based on the load information and the basic equipment information, and a centralized voltage control strategy is obtained according to the centralized voltage control strategy model, including: Based on the load information and basic equipment information, the distribution network demand response model, energy storage operation model and reactive power regulation equipment model are constructed respectively; Determine the control variables of the centralized voltage control strategy model based on the established distribution network demand response model, energy storage operation model and reactive power regulation equipment model; Taking the minimum network loss, minimum voltage deviation and minimum economic cost as the objective function, a centralized voltage control strategy model is constructed in combination with the determined control variables.

[0009] Furthermore, step S2 constructs a centralized voltage control strategy model based on load information and basic equipment information, and obtains a centralized voltage control strategy according to the centralized voltage control strategy model, and also includes converting the centralized voltage control strategy model into a mixed integer second-order cone programming model, and solving the mixed integer second-order cone programming model to obtain the centralized voltage control strategy.

[0010] Furthermore, in step S4, the real-time voltage regulating device is determined according to the basic device information, a real-time voltage control strategy model is constructed based on the real-time voltage regulating device and the operation status information, and the real-time voltage control strategy is obtained according to the real-time voltage control strategy model, including: Based on the device type and device access node, obtain the real-time voltage regulation device involved in real-time voltage control in the basic device information, and retrieve the operating status information of the active distribution network; The constraints of the real-time voltage control strategy model are constructed according to the real-time voltage regulation equipment and operation status information, and the real-time voltage control strategy model is constructed with minimizing the system node voltage deviation and distribution network loss as the objective function.

[0011] Further, the step S4 includes determining the real-time voltage regulation device according to the basic device information, building a real-time voltage control strategy model based on the real-time voltage regulation device and the operation status information, and obtaining the real-time voltage control strategy according to the real-time voltage control strategy model, and further includes: The real-time voltage regulating device participating in the real-time voltage control is taken as an independent intelligent agent in the Markov game algorithm, and a Markov game model is established according to the constructed real-time voltage control strategy model; The Markov game model is solved by a double-delayed deep deterministic policy gradient algorithm to obtain the real-time voltage control strategy.

[0012] Furthermore, the voltage regulation deviation information of the active distribution network in the corresponding feedback regulation period is determined according to the control result of the real-time voltage control strategy in step S6, including: According to the time period corresponding to the current feedback regulation cycle, the voltage regulation target value and load regulation information of the current feedback regulation cycle are matched with the centralized voltage control strategy; Determine a voltage final value according to a control result of a real-time voltage control strategy, and determine a first voltage deviation value according to a voltage regulation target value and the voltage final value; Determine the dispatching load and its load regulation value participating in the centralized voltage control according to the load regulation information, and determine the real-time voltage control device and its access node information participating in the real-time voltage control according to the real-time voltage control strategy; Associating the dispatching load with the real-time voltage control device according to the access node information of each real-time voltage control device participating in the real-time voltage control; Obtaining voltage regulation information of each real-time voltage control device in a current feedback regulation cycle, and determining a voltage regulation deviation value of each dispatching load according to a load regulation value of each dispatching load and voltage regulation information of the real-time voltage control device associated therewith; Voltage regulation deviation information is determined according to the first voltage deviation value and the voltage regulation deviation value of each dispatching load.

[0013] The influence of the real-time voltage control strategy on the centralized voltage control strategy is determined by the voltage deviation value, thereby providing an optimization basis for the subsequent optimization of the centralized voltage control strategy.

[0014] Furthermore, the optimization direction of the centralized voltage control strategy based on the voltage regulation deviation information matching in step S6 includes: Determine the voltage deviation value of the execution object corresponding to the centralized voltage control strategy according to the voltage regulation deviation information, and sort the execution objects in descending order according to the corresponding voltage deviation value; The execution objects are divided into types, and for each type of execution object, a weighted sum is performed according to the ranking result of the corresponding execution object and the voltage deviation value; The execution object type with the largest weighted sum result is selected, and the optimization direction of the centralized voltage control strategy is matched according to the execution object type.

[0015] Considering the large number of devices in the active distribution network, before optimizing the centralized voltage control strategy, the corresponding optimization direction is matched first, and the optimization object is screened by the optimization direction to reduce the optimization calculation amount and improve the optimization efficiency of the centralized voltage control strategy.

[0016] Furthermore, the centralized voltage control strategy is optimized according to the optimization direction of the centralized voltage control strategy obtained by matching and the voltage regulation deviation information of the active distribution network in the corresponding feedback regulation period in step S7, including: Selecting an execution object of the centralized voltage control strategy to be optimized according to the optimization direction of the centralized voltage control strategy obtained by matching; Determine the target voltage regulation value and the completed voltage regulation value of each corresponding execution object according to the centralized voltage control strategy, and determine the target voltage regulation deviation value to be optimized in combination with the voltage regulation deviation information; The screened execution objects are taken as optimization objects, the target voltage adjustment deviation value to be optimized is taken as the optimization target, and the voltage control strategy corresponding to each screened execution object in the centralized voltage control strategy is adjusted.

[0017] By adjusting the voltage control strategy of the execution object in the centralized voltage control strategy, coordination and synchronization between the execution objects of the real-time voltage control strategy and the centralized voltage control strategy are achieved, avoiding conflicts between the real-time voltage control strategy and the centralized voltage control strategy.

[0018] Furthermore, in step S5, the load change information and the system stability index are obtained based on the operation status information, and the feedback adjustment cycle is set based on the load change information and the system stability index, including: Setting a standard cycle length of a feedback regulation cycle, and taking the feedback regulation cycle of the standard cycle length as an initial feedback regulation cycle for optimization of a centralized voltage control strategy; Selecting a standard load, matching load data of the standard load from the operating status information, and determining a load change rate of the standard load according to the matched load data; Set the evaluation criteria for system stability indicators and obtain system stability indicators in combination with the operation status information; Match the load change rate of the standard load and the system stability index with the corresponding adjustment conditions, and determine the corresponding adjustment rules according to the matched adjustment conditions; The standard cycle length is used as the adjustment basis, and the cycle length of the next feedback adjustment cycle is determined according to the corresponding adjustment rules.

[0019] The setting of the feedback regulation cycle has an important impact on the regulation effect. If the cycle is too long, it may lead to a large voltage deviation; if the cycle is too short, it may increase the system burden and introduce unnecessary control fluctuations. Therefore, load changes and system stability are selected as the basis for feedback cycle regulation, and the cycle length of the feedback regulation cycle is adjusted in real time to ensure that the setting of the cycle length can not only meet the real-time requirements of voltage regulation, but also avoid unnecessary system burden and cost increase.

[0020] The beneficial effects of the present invention are: On the basis of formulating a long-time centralized voltage control strategy in combination with load dispatch and reactive equipment, a short-time real-time voltage control strategy is formulated. In the process of implementing the centralized voltage control strategy, the real-time voltage regulation device is used to adjust the short-time voltage fluctuations in real time according to the real-time voltage control strategy, and the centralized control strategy is periodically optimized according to the control results of the real-time voltage control strategy to achieve coordinated synchronization between the long-time centralized voltage control strategy and the short-time real-time voltage control strategy, avoid conflicts between the control results of the two control strategies, and ensure the stable operation of the active distribution network.

[0021] Furthermore, load changes and system stability are selected as the basis for feedback cycle regulation, and the cycle length of the feedback regulation cycle is adjusted in real time to ensure that the setting of the cycle length can not only meet the real-time requirements of voltage regulation, but also avoid unnecessary system burden and cost increase.

[0022] Considering the large number of devices in the active distribution network, before optimizing the centralized voltage control strategy, the corresponding optimization direction is matched first, and the optimization object is screened by the optimization direction to reduce the optimization calculation amount and improve the optimization efficiency of the centralized voltage control strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a schematic diagram of a process of the present invention. DETAILED DESCRIPTION

[0024] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0025] Example:

[0026] Active distribution network voltage control method based on multiple time scales, such as Figure 1 As shown, including: Step S1, obtaining load information and basic equipment information of the active distribution network; Step S2, constructing a centralized voltage control strategy model based on load information and basic equipment information, and obtaining a centralized voltage control strategy according to the centralized voltage control strategy model; Step S3, executing a centralized voltage control strategy and collecting operating status information of the active distribution network in real time; Step S4, determining the real-time voltage regulation device according to the basic device information, building a real-time voltage control strategy model based on the real-time voltage regulation device and the operation status information, obtaining the real-time voltage control strategy according to the real-time voltage control strategy model, and executing the real-time voltage control strategy; Step S5, obtaining load change information and system stability indicators based on the operating status information, setting a feedback adjustment cycle based on the load change information and the system stability indicators, and obtaining a control result of a corresponding real-time voltage control strategy at the end of the current feedback adjustment cycle; Step S6, determining the voltage regulation deviation information of the active distribution network in the corresponding feedback regulation period according to the control result of the real-time voltage control strategy, and matching the optimization direction of the centralized voltage control strategy based on the voltage regulation deviation information; Step S7, optimizing the centralized voltage control strategy according to the optimization direction of the centralized voltage control strategy obtained by matching and the voltage regulation deviation information of the active distribution network in the corresponding feedback regulation cycle, and executing the optimized centralized voltage control strategy at the beginning of the next feedback regulation cycle until the centralized voltage control strategy is completed.

[0027] In view of the high complexity and polymorphic operation of the active distribution network, its regulation capability is further developed from the demand side. Both demand response and energy storage systems are used as the execution objects of the centralized voltage control strategy. The adjustability of electricity consumption behavior and charging and discharging behavior is utilized. The voltage of the active distribution network is simply adjusted through load regulation, enriching the voltage regulation methods of the active distribution network.

[0028] Since the centralized voltage control strategy is a long-time voltage regulation method, it is difficult to handle short-time voltage fluctuations in the active distribution network. Therefore, while executing the centralized voltage control strategy, it is combined with short-time real-time voltage regulation for coordinated voltage regulation. The short-time real-time voltage control strategy is used to timely detect and handle sudden voltage fluctuations.

[0029] Although the execution objects and formulation basis of the real-time voltage control strategy and the centralized voltage control strategy are not consistent, there are correlations and similarities between the execution objects. In order to avoid conflicts between the control results of the two control strategies, the centralized voltage control strategy is adjusted in stages according to the voltage control situation of the real-time voltage control strategy to maintain coordinated synchronization between the execution objects of the real-time voltage control strategy and the execution objects of the centralized voltage control strategy to ensure the stable operation of the active distribution network.

[0030] The centralized voltage control strategy model is constructed based on the load information and the basic equipment information, and the centralized voltage control strategy is obtained according to the centralized voltage control strategy model, including: Based on the load information and basic equipment information, the distribution network demand response model, energy storage operation model and reactive power regulation equipment model are constructed respectively; Determine the control variables of the centralized voltage control strategy model based on the established distribution network demand response model, energy storage operation model and reactive power regulation equipment model; Taking the minimum network loss, minimum voltage deviation and minimum economic cost as the objective function, a centralized voltage control strategy model is constructed in combination with the determined control variables.

[0031] Among them, before establishing the distribution network demand response model, the flexible load participating in the active distribution network demand response is obtained according to the load information, including the transferable load and the curtailable load, and then the distribution network demand response model is constructed according to the load dispatching potential of the corresponding flexible load.

[0032] The expression of the distribution network demand response model is: ; Where D refers to the active distribution network, T is the set dispatching period, and L is one of the flexible loads. and are the load powers transferred in and out by the flexible load L during period t, and They represent the maximum loads transferred in and out of the flexible load L during period t, is the load power that can be reduced by the flexible load L during period t, It represents the maximum load power that can be reduced by the flexible load L during period t; Represents the dispatchable interval of transferable load.

[0033] Combined with the above distribution network demand response model, the corresponding load response supplementary cost function is constructed, and its expression is: ; in, represents the flexible load set in the active distribution network D; and They are the compensation prices for transferring out unit power and reducing unit power of flexible load L respectively.

[0034] Before establishing the energy storage operation model, the energy storage system configured in the active distribution network is determined according to the basic equipment information, and the energy storage operation model is constructed based on the operation status of the energy storage system. The expression is: ; in, and are the state of charge of the energy storage system at node i in the active distribution network at time t and the state of charge of the energy storage system at node i in the active distribution network at time t+1, To optimize the time period; and is the charging and discharging efficiency of the energy storage system; and are the storage capacities before and after the energy storage system optimization; and is the charging and discharging power of the energy storage system at node i in the active distribution network at time t, is the maximum output power of the energy storage system, and are the minimum and maximum state of charge of the energy storage system at node i in the active distribution network, respectively.

[0035] Combined with the above energy storage operation model, the corresponding energy storage system investment cost function is constructed, and its expression is: ; in, is the overall cost of the energy storage system, and are the initial investment cost and operation and maintenance cost of the energy storage system, is the investment discount factor, k is the discount rate, y is the life cycle of the energy storage system, is the unit capacity investment cost of energy storage, is the operation and maintenance cost equivalence coefficient of the energy storage system.

[0036] The reactive power regulation equipment model specifically includes an on-load tap-changing transformer equipment model and a group switching capacitor model.

[0037] Specifically, the expression of the on-load tap-changing transformer equipment model is: ; in, and are the voltage amplitudes of the nodes at the high-voltage side and low-voltage side of the on-load tap-changing transformer at time t respectively; is the square of the transformation ratio of the on-load tap-changing transformer at time t; and The upper and lower limits of the square of the on-load tap-changing transformer ratio; It is the difference between the square of the transformation ratio of the on-load tap-changing transformer in gear s and gear s-1; Is a 0-1 identifier.

[0038] The expression of the group switching capacitor model is: ; in, is the number of grouped switching capacitors put into operation at node j of the active distribution network at time t; The reactive power compensated by each group switching capacitor.

[0039] Based on the above distribution network demand response model, energy storage operation model and reactive power regulation equipment model, the centralized voltage control strategy model is constructed with the minimum network loss, minimum voltage deviation and minimum economic cost as the objective function, considering the control variables such as demand response scheduling, energy storage system charging and discharging, capacitor switching and on-load tap-changing transformer gear adjustment. The expression of its objective function is: ; Where F is the objective function value of the centralized voltage control strategy model, , , are the weight coefficients corresponding to network loss, voltage deviation and economic cost of distribution network, , and are the network loss, voltage deviation and economic cost of the active distribution network, respectively; N is the total number of nodes in the active distribution network, is the loss coefficient between node i and node j, and are the node voltages at node i and node j in the active distribution network at time t, is the phase difference between the voltages at node i and node j at time t, is the rated node voltage of the active distribution network.

[0040] The method of constructing a centralized voltage control strategy model based on load information and basic equipment information, and obtaining a centralized voltage control strategy according to the centralized voltage control strategy model, also includes converting the centralized voltage control strategy model into a mixed integer second-order cone programming model, and solving the mixed integer second-order cone programming model to obtain the centralized voltage control strategy.

[0041] Specifically, the constraints corresponding to the centralized voltage control strategy are obtained, such as voltage constraints, reactive power balance constraints, reactive supplementary equipment constraints, power flow constraints, and integer variable constraints, and the above nonlinear constraints, such as the voltage square term, the relationship between reactive power, voltage, and current, are converted into second-order cone constraints through second-order cone relaxation. Then, a complete mixed integer second-order cone programming model is formed by combining the objective function of the corresponding centralized voltage control strategy and all constraints.

[0042] Then, the YALMIP tool is used to call solvers such as MOSEK and CPLEX, and the mixed integer second-order cone programming model is solved in combination with the GUROBI tool to obtain the optimal solution that meets all constraints or reaches the termination conditions set by the solver, such as the number of iterations.

[0043] Finally, according to the solution results output by the solver, the demand response strategy, energy storage system charging and discharging strategy, reactive power regulation equipment switching state control strategy, etc. are determined to obtain the centralized voltage control strategy.

[0044] The method of determining a real-time voltage regulating device according to basic device information, constructing a real-time voltage control strategy model based on the real-time voltage regulating device and operation status information, and obtaining a real-time voltage control strategy according to the real-time voltage control strategy model includes: Based on the device type and device access node, obtain the real-time voltage regulation device involved in real-time voltage control in the basic device information, and retrieve the operating status information of the active distribution network; The constraints of the real-time voltage control strategy model are constructed based on the real-time voltage regulation equipment and operation status information.

[0045] The real-time voltage regulation equipment includes a photovoltaic inverter and a static VAR compensator SVC. Therefore, the equipment capable of real-time voltage regulation is first screened out according to the equipment type, and then it is determined whether it overlaps with the control area corresponding to the centralized voltage control strategy according to the location of its access node, and the equipment capable of real-time voltage regulation that overlaps with the control area corresponding to the centralized voltage control strategy is selected as the execution object of the real-time voltage control strategy.

[0046] Based on the equipment that can perform real-time voltage regulation, and taking minimizing the system node voltage deviation and distribution network loss as the objective function, a real-time voltage control strategy model is constructed. The specific objective function is: ; in, and are the weight coefficients of system node voltage deviation and distribution network loss respectively.

[0047] The constraints include power flow constraints of distribution network nodes, reactive power compensation constraints of reactive power compensators, and active power output constraints and reactive power output constraints of photovoltaic inverters.

[0048] Among them, the expression of the power flow constraint of the distribution network node is: ; in, and is the active output and active load of the PV inverter at node i at time t; and are the real and imaginary parts of the admittance of the branch formed by node i and node j respectively; , as well as They are the reactive output of the photovoltaic inverter at node i at time t, the reactive compensation of the reactive compensator and the reactive load demand.

[0049] The expression of reactive power compensation constraint of reactive power compensator is: ; in, and They are the upper and lower limits of reactive power compensation of the reactive power compensator respectively.

[0050] The expression of active output constraint of photovoltaic inverter is: ; in, is the active output rating of the PV inverter at node i.

[0051] The expression of reactive output constraint of photovoltaic inverter is: ; in, Represents the apparent capacity of the PV inverter at node i.

[0052] The method further includes determining the real-time voltage regulating device according to the basic device information, constructing a real-time voltage control strategy model based on the real-time voltage regulating device and the operation status information, and obtaining the real-time voltage control strategy according to the real-time voltage control strategy model. The real-time voltage regulating device participating in the real-time voltage control is taken as an independent intelligent agent in the Markov game algorithm, and a Markov game model is established according to the constructed real-time voltage control strategy model; The Markov game model is solved by a double-delayed deep deterministic policy gradient algorithm to obtain the real-time voltage control strategy.

[0053] Markov game is a mathematical framework used to describe multi-agent decision-making problems. In a Markov game, the decision of the agent depends on the current state, and the state transition has the Markov property, that is, the future state depends only on the current state and the action of the agent, but has nothing to do with the past state. In a zero-sum game, the goal of the agent is to maximize its own benefits and reduce the opponent's benefits accordingly.

[0054] In this embodiment, the real-time voltage regulation device involved in real-time voltage control is used as an independent agent in the Markov game algorithm, and the state observation of the agent p at the i-th node at the t-th time is defined as , including the load active power of node i at time t , reactive power , Photovoltaic inverter active output , voltage amplitude and reactive load demand Define the action of agent p at node i at time t as , including the reactive power output of the photovoltaic inverter of node i at time t And reactive power compensation output of reactive power compensator . Thus, a Markov game model is established based on the constructed real-time voltage control strategy model.

[0055] The double-delayed deep deterministic policy gradient algorithm is a policy gradient algorithm improved on the basis of deep deterministic policy gradient, which is used to solve the optimal strategy in the Markov decision process.

[0056] When using the double-delayed deep deterministic policy gradient algorithm to solve the Markov game model, it is necessary to first define the game environment, clarify the state space, action space, payoff function and transition probability of the game, and then initialize the network parameters of the two critic networks, one actor network and the corresponding target network. The state observations and actions of the intelligent agents defined above are stored in the experience pool to determine the experience data.

[0057] And iterative training is performed based on empirical data. Take one of the trainings as an example: A batch of data is randomly sampled from the experience pool, the Q value is estimated using two critic networks respectively, and the smaller one is selected as the target Q value, the loss function of the critic network is calculated, and the parameters of the critic network are updated. If necessary, the actor network is updated with a delay, the update frequency is reduced, and the target network is updated with a delay.

[0058] Through the iterative training process, the strategy of the intelligent agent is gradually improved to achieve the goal of maximizing benefits and obtain the optimal decision, that is, the real-time voltage control strategy.

[0059] In the above solution process, the double-delayed deep deterministic policy gradient algorithm also specifically adopts three methods: truncated double Q learning, target policy smoothing and delayed policy updating to reduce the bias and variance problems introduced by the function in the Markov decision process.

[0060] Truncated double Q learning is specifically as follows: the Q value is estimated by establishing two independent critic networks, and the smaller of the two Q values ​​is selected as the target Q value to avoid over-estimation of the Q value.

[0061] Target strategy smoothing is to add truncation noise to the target strategy to increase the generalization ability of the algorithm and alleviate the overfitting problem.

[0062] The specific method of delayed policy update is to reduce the update frequency of the actor network and wait as long as possible for the critic network training to converge before updating to reduce the accumulated error.

[0063] The voltage regulation deviation information of the active distribution network in the corresponding feedback regulation period is determined according to the control result of the real-time voltage control strategy, including: According to the time period corresponding to the current feedback regulation cycle, the voltage regulation target value and load regulation information of the current feedback regulation cycle are matched with the centralized voltage control strategy; Determine a voltage final value according to a control result of a real-time voltage control strategy, and determine a first voltage deviation value according to a voltage regulation target value and the voltage final value; Determine the dispatching load and its load regulation value participating in the centralized voltage control according to the load regulation information, and determine the real-time voltage control device and its access node information participating in the real-time voltage control according to the real-time voltage control strategy; Associating the dispatching load with the real-time voltage control device according to the access node information of each real-time voltage control device participating in the real-time voltage control; Obtaining voltage regulation information of each real-time voltage control device in a current feedback regulation cycle, and determining a voltage regulation deviation value of each dispatching load according to a load regulation value of each dispatching load and voltage regulation information of the real-time voltage control device associated therewith; Voltage regulation deviation information is determined according to the first voltage deviation value and the voltage regulation deviation value of each dispatching load.

[0064] Since the execution object of the real-time voltage control strategy is inconsistent with the execution object of the centralized voltage control strategy, but the control objects are both the voltage of the active distribution network, after the real-time voltage control strategy is executed, if the centralized voltage control strategy continues to regulate the voltage according to the original strategy, problems such as over-regulation and voltage regulation conflict may occur.

[0065] In particular, the execution objects of the centralized voltage control strategy include the demand side that indirectly realizes voltage regulation through load scheduling and the energy storage system that indirectly realizes voltage regulation through charging and discharging behavior. If the corresponding control strategy cannot be adjusted in time, the load scheduling and charging and discharging behavior will cause voltage fluctuations in the active distribution network.

[0066] Therefore, the influence of the real-time voltage control strategy on the execution of the centralized voltage control strategy is judged according to the control result of the real-time voltage control strategy, so as to adjust the centralized voltage control strategy.

[0067] In order to quantify the impact of the control results of the real-time voltage control strategy on the execution of the centralized voltage control strategy, the voltage is used as the quantification basis. The centralized voltage control strategy is first matched according to the corresponding time period of the current feedback adjustment cycle, so as to determine the voltage regulation target value and load regulation information originally planned by the centralized voltage control strategy when the real-time voltage control strategy is not executed.

[0068] The voltage regulation target value includes the voltage value of the active distribution network displayed at the end of the feedback regulation cycle after all execution objects operate according to the corresponding control strategy content.

[0069] The load regulation information includes demand response and load scheduling operation information and charging and discharging behavior regulation operation information made by the energy storage system for voltage regulation.

[0070] Then, the final voltage value is determined based on the control result of the real-time voltage control strategy, that is, the voltage value actually displayed by the current active distribution network after the real-time voltage control strategy intervenes. By comparing the final voltage value and the voltage regulation target value, the impact on the execution effect of the centralized voltage control strategy after the real-time voltage control strategy intervenes can be determined, and the impact can be quantified as the first voltage deviation value, providing data support for the subsequent optimization of the centralized voltage control strategy.

[0071] Since the centralized voltage control strategy includes three execution objects: demand response, energy storage and reactive power regulation equipment, each execution object is affected to different degrees by the real-time voltage control equipment due to its adjustment method, access location, etc. Therefore, the voltage deviation value of each execution object under the influence of the real-time voltage control strategy is obtained to provide data support for the subsequent selection of optimization direction.

[0072] Among them, although the reactive power regulation equipment and the real-time voltage control equipment belong to different types of equipment, they both regulate voltage through reactive power compensation. Their voltage regulation is less affected by the operation of the real-time voltage control equipment. Therefore, voltage deviation analysis is only performed on demand response and energy storage with load regulation.

[0073] Since the load dispatching effects of demand response and energy storage are affected by grid voltage fluctuations, when real-time voltage control equipment intervenes, the voltage regulation effects of demand response and energy storage will deviate from expectations.

[0074] According to the real-time voltage control strategy, the real-time voltage control devices and their access node information involved in the real-time voltage control are determined. The real-time voltage control strategy realizes voltage regulation on the spot. Therefore, the influence range of the voltage regulation can be determined according to the access node of the real-time voltage control device, thereby determining the dispatching load affected by its operation and determining the association between the dispatching load and the real-time voltage control device. Among them, the association relationship can be obtained through association analysis methods such as grey association degree.

[0075] The voltage regulation information of each real-time voltage control device in the current feedback regulation cycle is obtained to determine the voltage fluctuation caused by it. Then, through the association between the dispatched load and the real-time voltage control device, the voltage regulation value finally displayed when the corresponding load regulation value is adjusted under the influence of all the associated real-time voltage control devices on the dispatched load is determined.

[0076] Then, the originally planned voltage regulation value is determined according to the load regulation value, and the difference between the originally planned voltage regulation value and the calculated voltage regulation value under the influence of the real-time voltage control device is calculated to determine the voltage regulation deviation value.

[0077] The optimization direction of the centralized voltage control strategy based on voltage regulation deviation information matching includes: Determine the voltage deviation value of the execution object corresponding to the centralized voltage control strategy according to the voltage regulation deviation information, and sort the execution objects in descending order according to the corresponding voltage deviation value; The execution objects are divided into types, and for each type of execution object, a weighted sum is performed according to the ranking result of the corresponding execution object and the voltage deviation value; The execution object type with the largest weighted sum result is selected, and the optimization direction of the centralized voltage control strategy is matched according to the execution object type.

[0078] In order to improve the optimization efficiency and reduce the amount of calculation, the optimization direction is first determined according to the voltage regulation information.

[0079] Among them, since the voltage regulation deviation information reflects the voltage regulation deviation value of each execution object affected by the real-time voltage control device in the centralized voltage control strategy, the degree of influence of these execution objects on the real-time voltage control device can be evaluated by sorting order and weighted sum of corresponding voltage deviation values.

[0080] The execution object type most affected by the real-time voltage control device is selected, and based on this, the corresponding optimization direction is determined.

[0081] Specific optimization directions include various combinations of three regulation operations: adjusting the demand response strategy, adjusting the charging and discharging behavior of the energy storage system, and adjusting the reactive power injected by the reactive regulation equipment.

[0082] If the demand response strategy is most affected by the real-time voltage control device and causes the largest voltage deviation, then the optimization direction is to adjust the charging and discharging behavior of the energy storage system and the reactive power injected by the reactive power regulation device to correct the voltage deviation. This can correct the voltage deviation caused by the execution of the real-time voltage control strategy and avoid further expansion of the impact of subsequent real-time voltage control strategies.

[0083] The centralized voltage control strategy is optimized according to the optimization direction of the centralized voltage control strategy obtained by matching and the voltage regulation deviation information of the active distribution network in the corresponding feedback regulation period, including: Selecting an execution object of the centralized voltage control strategy to be optimized according to the optimization direction of the centralized voltage control strategy obtained by matching; Determine the target voltage regulation value and the completed voltage regulation value of each corresponding execution object according to the centralized voltage control strategy, and determine the target voltage regulation deviation value to be optimized in combination with the voltage regulation deviation information; The screened execution objects are taken as optimization objects, the target voltage adjustment deviation value to be optimized is taken as the optimization target, and the voltage control strategy corresponding to each screened execution object in the centralized voltage control strategy is adjusted.

[0084] After determining the optimization direction, the execution objects that need to be adjusted in the strategy content can be determined according to the optimization direction. For example, when the optimization direction is to adjust the charging and discharging behavior of the energy storage system and the reactive power injected by the reactive regulation device to correct the voltage deviation, the energy storage device and the reactive regulation device are selected as the execution objects.

[0085] When selecting execution objects, it is necessary to evaluate the adjustment capability of each execution object, and only optimize the policy content for execution objects with adjustment margin.

[0086] Since the execution cycle of the centralized voltage control strategy must be longer than the feedback regulation cycle, when optimizing the centralized voltage control strategy, it must be in the control process. Therefore, the control strategy content it adjusts must be the strategy content of the time period after the feedback regulation cycle ends. Therefore, the target voltage regulation value of each corresponding execution object and the completed voltage regulation value are determined, and the original voltage regulation values ​​of all execution objects screened out in the centralized voltage control strategy are determined.

[0087] Then, the first voltage deviation value that needs to be compensated and corrected is determined according to the voltage regulation deviation information, and the target voltage regulation deviation value to be optimized is determined in combination with the corresponding original voltage regulation value, that is, the voltage regulation value after correcting the first voltage deviation value.

[0088] The screened execution objects are taken as optimization objects, the target voltage regulation deviation value to be optimized is taken as optimization target, and the voltage control strategy corresponding to each screened execution object in the centralized voltage control strategy is adjusted. Specifically, it can be obtained by iteration through optimization algorithms such as genetic algorithm and particle swarm optimization algorithm.

[0089] Moreover, this optimization adjustment needs to be based on the adjustment margin and adjustment cost of each screened execution object as constraints to avoid affecting the operational safety of the execution object.

[0090] The step S5 of acquiring load change information and system stability index based on the operation status information and setting a feedback adjustment cycle based on the load change information and the system stability index includes: Setting a standard cycle length of a feedback regulation cycle, and taking the feedback regulation cycle of the standard cycle length as an initial feedback regulation cycle for optimization of a centralized voltage control strategy; Selecting a standard load, matching load data of the standard load from the operating status information, and determining a load change rate of the standard load according to the matched load data; Set the evaluation criteria for system stability indicators and obtain system stability indicators in combination with the operation status information; Match the load change rate of the standard load and the system stability index with the corresponding adjustment conditions, and determine the corresponding adjustment rules according to the matched adjustment conditions; The standard cycle length is used as the adjustment basis, and the cycle length of the next feedback adjustment cycle is determined according to the corresponding adjustment rules.

[0091] The setting of the feedback regulation cycle has an important impact on the regulation effect. If the cycle is too long, it may lead to a large voltage deviation. If the cycle is too short, it may increase the system burden and introduce unnecessary control fluctuations. In addition, the feedback regulation is based on the completed historical data and will not affect the subsequent strategy execution. Therefore, the cycle length of the feedback regulation cycle is adjusted according to the execution of the centralized voltage control strategy and the real-time voltage control strategy to further ensure the stable operation of the active distribution network.

[0092] At the same time, since the centralized voltage control strategy involves demand response and energy storage system regulation, both have significant time characteristics, which means that the regulation effects they exhibit are different in different time periods. In order to achieve coordination and synchronization between the centralized voltage control strategy with time characteristics and the real-time voltage control strategy with real-time regulation, it is even more necessary to adaptively adjust the length of the feedback regulation cycle to ensure the coordination and synchronization effect.

[0093] Among them, load change information and system stability index are important indicators for judging and evaluating the operating status of the active distribution network, which can intuitively reflect the regulation effect of the centralized voltage control strategy and the real-time voltage control strategy. Therefore, load change information and system stability index are selected as the basis for the adjustment of the feedback regulation cycle.

[0094] The load change information can be obtained by selecting representative loads such as key loads as standard loads and determining their corresponding load change rates, thereby reflecting the operation status of the active distribution network.

[0095] The system stability index can be calculated through operating status information such as frequency deviation and power factor to determine the system stability margin under the regulation of centralized voltage control strategy and real-time voltage control strategy.

[0096] The load change rate and the system stability index can be used to determine the regulation conditions that reflect the operation status of the active distribution network. Therefore, the load change rate and the system stability index are matched with the corresponding regulation conditions.

[0097] For example, when the load change rate is low and the system stability index is high, the regulation conditions for the stable operation of the active distribution network can be matched, and then the corresponding regulation rules can be determined based on the regulation conditions for the stable operation of the active distribution network, that is, the standard cycle length is used as the regulation basis and the cycle length is extended.

[0098] If the load change rate is high or the system stability index is low, the regulation conditions of the active distribution network with operating fluctuations can be matched. Then, the corresponding regulation rules are determined according to the regulation conditions of the active distribution network with operating fluctuations, that is, the standard cycle length is used as the regulation basis, and the cycle length is shortened. The length of the cycle can be extended or shortened by adjusting the number of times and the frequency.

[0099] Ensure that the cycle length setting can not only meet the real-time requirements of voltage regulation, but also avoid unnecessary system burden and cost increase.

[0100] Taking half an hour as the standard cycle length of the feedback regulation cycle as an example, the operation information within half an hour after the centralized voltage control strategy is started is collected in real time.

[0101] Based on the collected operating information, if it is determined that there is a high load change rate or a low system stability index within the half hour, it is determined that the length of the feedback adjustment cycle needs to be shortened. Since the feedback adjustment has just started, the original adjustment rule is directly used at this time, that is, the half-time method is used to adjust the next feedback adjustment cycle to 15 minutes.

[0102] At the end of the next feedback regulation cycle, that is, 45 minutes after the centralized voltage control strategy is started, the regulation conditions are determined based on the operating information in the next 15 minutes. If the load change rate is still high or the system stability index is low within the 15 minutes, the original regulation rule, that is, half of the original regulation rule, is subtracted by 1 minute, and the cycle length of the feedback regulation cycle is adjusted to 14 minutes.

[0103] Similarly, if a high load change rate or a low system stability index occurs continuously in the future, the length of the next feedback regulation cycle will be determined based on the number of feedback regulation cycles in which this situation occurs and the corresponding number of minutes will be subtracted from the original regulation rules.

[0104] If it is determined that the load change rate is low or the system stability index is high within this half hour, it is determined that the length of the feedback adjustment cycle needs to be increased. Since the feedback adjustment has just started, the original adjustment rule is directly used at this time, that is, increasing the cycle length by 15 minutes, and the next feedback adjustment cycle is adjusted to 45 minutes.

[0105] At the end of the next feedback regulation cycle, that is, 45 minutes after the centralized voltage control strategy begins, the regulation conditions are determined based on the operating information in the next 15 minutes. If the load change rate is still low or the system stability index is high within the 15 minutes, then based on the original regulation rule, that is, increasing the cycle length by 15 minutes, an additional 10 minutes is added to adjust the cycle length of the feedback regulation cycle to 55 minutes.

[0106] Similarly, if the load change rate is low or the system stability index is high in the subsequent period, the length of the next feedback regulation cycle will be determined based on the number of feedback regulation cycles in which this situation occurs and the original regulation rules. The cycle length will be increased by a corresponding number of times in units of 10 minutes to determine the cycle length of the next feedback regulation cycle.

[0107] Furthermore, since the purpose of setting the feedback regulation cycle is to coordinate the centralized voltage control strategy and the real-time voltage control strategy, the regulation setting of the feedback regulation cycle can be stopped after the execution of the centralized voltage control strategy is completed.

[0108] The above-described embodiment is only a preferred solution of the present invention and does not limit the present invention in any form. There are other variations and modifications without exceeding the technical solution described in the claims.

Claims

1. An active distribution network voltage control method based on multiple time scales, characterized in that: include: Step S1, obtaining load information and basic equipment information of the active distribution network; Step S2, constructing a centralized voltage control strategy model based on load information and basic equipment information, and obtaining a centralized voltage control strategy according to the centralized voltage control strategy model; Step S3, executing a centralized voltage control strategy and collecting operating status information of the active distribution network in real time; Step S4, determining the real-time voltage regulation device according to the basic device information, building a real-time voltage control strategy model based on the real-time voltage regulation device and the operation status information, obtaining the real-time voltage control strategy according to the real-time voltage control strategy model, and executing the real-time voltage control strategy; Step S5, obtaining load change information and system stability indicators based on the operating status information, setting a feedback adjustment cycle based on the load change information and the system stability indicators, and obtaining a control result of a corresponding real-time voltage control strategy at the end of the current feedback adjustment cycle; Step S6, determining the voltage regulation deviation information of the active distribution network in the corresponding feedback regulation period according to the control result of the real-time voltage control strategy, and matching the optimization direction of the centralized voltage control strategy based on the voltage regulation deviation information; Step S7, optimizing the centralized voltage control strategy according to the optimization direction of the centralized voltage control strategy obtained by matching and the voltage regulation deviation information of the active distribution network in the corresponding feedback regulation cycle, and executing the optimized centralized voltage control strategy at the beginning of the next feedback regulation cycle until the centralized voltage control strategy is completed.

2. The active distribution network voltage control method based on multiple time scales according to claim 1 is characterized in that: In step S2, a centralized voltage control strategy model is constructed based on load information and basic equipment information, and a centralized voltage control strategy is obtained according to the centralized voltage control strategy model, including: Based on the load information and basic equipment information, the distribution network demand response model, energy storage operation model and reactive power regulation equipment model are constructed respectively; Determine the control variables of the centralized voltage control strategy model based on the established distribution network demand response model, energy storage operation model and reactive power regulation equipment model; Taking the minimum network loss, minimum voltage deviation and minimum economic cost as the objective function, a centralized voltage control strategy model is constructed in combination with the determined control variables.

3. The active distribution network voltage control method based on multiple time scales according to claim 2 is characterized in that: The step S2 described in constructing a centralized voltage control strategy model based on load information and basic equipment information, and obtaining a centralized voltage control strategy according to the centralized voltage control strategy model also includes converting the centralized voltage control strategy model into a mixed integer second-order cone programming model, and solving the mixed integer second-order cone programming model to obtain the centralized voltage control strategy.

4. The active distribution network voltage control method based on multiple time scales according to claim 1 is characterized in that: The step S4 includes determining the real-time voltage regulating device according to the basic device information, building a real-time voltage control strategy model based on the real-time voltage regulating device and the operation status information, and obtaining the real-time voltage control strategy according to the real-time voltage control strategy model, including: Based on the device type and device access node, obtain the real-time voltage regulation device involved in real-time voltage control in the basic device information, and retrieve the operating status information of the active distribution network; The constraints of the real-time voltage control strategy model are constructed according to the real-time voltage regulation equipment and operation status information, and the real-time voltage control strategy model is constructed with minimizing the system node voltage deviation and distribution network loss as the objective function.

5. The active distribution network voltage control method based on multiple time scales according to claim 4 is characterized in that: The step S4 includes determining the real-time voltage regulating device according to the basic device information, building a real-time voltage control strategy model based on the real-time voltage regulating device and the operation status information, and obtaining the real-time voltage control strategy according to the real-time voltage control strategy model, and further comprising: The real-time voltage regulating device participating in the real-time voltage control is taken as an independent intelligent agent in the Markov game algorithm, and a Markov game model is established according to the constructed real-time voltage control strategy model; The Markov game model is solved by a double-delayed deep deterministic policy gradient algorithm to obtain the real-time voltage control strategy.

6. The active distribution network voltage control method based on multiple time scales according to claim 1 is characterized in that: Determining the voltage regulation deviation information of the active distribution network in the corresponding feedback regulation period according to the control result of the real-time voltage control strategy in step S6 includes: According to the time period corresponding to the current feedback regulation cycle, the voltage regulation target value and load regulation information of the current feedback regulation cycle are matched with the centralized voltage control strategy; Determine a voltage final value according to a control result of a real-time voltage control strategy, and determine a first voltage deviation value according to a voltage regulation target value and the voltage final value; Determine the dispatching load and its load regulation value participating in the centralized voltage control according to the load regulation information, and determine the real-time voltage control device and its access node information participating in the real-time voltage control according to the real-time voltage control strategy; Associating the dispatching load with the real-time voltage control device according to the access node information of each real-time voltage control device participating in the real-time voltage control; Obtaining voltage regulation information of each real-time voltage control device in a current feedback regulation cycle, and determining a voltage regulation deviation value of each dispatching load according to a load regulation value of each dispatching load and voltage regulation information of the real-time voltage control device associated therewith; Voltage regulation deviation information is determined according to the first voltage deviation value and the voltage regulation deviation value of each dispatching load.

7. The active distribution network voltage control method based on multiple time scales according to claim 6 is characterized in that: The optimization direction of matching the centralized voltage control strategy based on the voltage regulation deviation information in step S6 includes: Determine the voltage deviation value of the execution object corresponding to the centralized voltage control strategy according to the voltage regulation deviation information, and sort the execution objects in descending order according to the corresponding voltage deviation value; The execution objects are divided into types, and for each type of execution object, a weighted sum is performed according to the ranking result of the corresponding execution object and the voltage deviation value; The execution object type with the largest weighted sum result is selected, and the optimization direction of the centralized voltage control strategy is matched according to the execution object type.

8. The active distribution network voltage control method based on multiple time scales according to claim 7 is characterized in that: The step S7 described in the invention optimizes the centralized voltage control strategy according to the optimization direction of the centralized voltage control strategy obtained by matching and the voltage regulation deviation information of the active distribution network in the corresponding feedback regulation period, including screening the execution object of the centralized voltage control strategy to be optimized according to the optimization direction of the centralized voltage control strategy obtained by matching; determining the target voltage regulation value and the completed voltage regulation value of each corresponding execution object according to the centralized voltage control strategy, and determining the target voltage regulation deviation value to be optimized in combination with the voltage regulation deviation information; taking the screened execution objects as the optimization objects, and the target voltage regulation deviation value to be optimized as the optimization target, and adjusting the voltage control strategy corresponding to each screened execution object in the centralized voltage control strategy.

9. The active distribution network voltage control method based on multiple time scales according to claim 1 is characterized in that: The step S5 of acquiring load change information and system stability index based on the operation status information and setting a feedback adjustment cycle based on the load change information and the system stability index includes: Setting a standard cycle length of a feedback regulation cycle, and taking the feedback regulation cycle of the standard cycle length as an initial feedback regulation cycle for optimization of a centralized voltage control strategy; Selecting a standard load, matching load data of the standard load from the operating status information, and determining a load change rate of the standard load according to the matched load data; Set the evaluation criteria for system stability indicators and obtain system stability indicators in combination with the operation status information; Match the load change rate of the standard load and the system stability index with the corresponding adjustment conditions, and determine the corresponding adjustment rules according to the matched adjustment conditions; The standard cycle length is used as the adjustment basis, and the cycle length of the next feedback adjustment cycle is determined according to the corresponding adjustment rules.

Citation Information

Patent Citations

  • Power system voltage control method based on near-end strategy optimization algorithm

    CN115409650A

  • Multi-integrated energy system cooperative attack toughness scheduling method considering integrated demand response based on state adversarial multi-agent deep reinforcement learning

    CN118229012A

  • Active power distribution network real-time voltage control method based on reactive power regulation of photovoltaic inverter

    CN118316135A

  • Method for multi-time scale voltage quality control based on reinforcement learning in a power distribution network

    US20220405633A1

Cited By

  • Voltage coordination control method and system based on distributed optical storage

    CN120749763A

  • Voltage coordination control method and system based on distributed optical storage

    CN120749763B

  • Power distribution network adaptive voltage regulation method and system based on real-time data feedback

    CN121440646A