A power distribution network voltage control method, system, device, medium and product

Abstract

The application discloses a power distribution network voltage control method, system, device, medium and product, relates to the field of power system automation, and comprises the following steps: acquiring real-time operation data of a power distribution network; performing real-time cluster division on nodes of the power distribution network according to the real-time operation data of the power distribution network; constructing a multi-objective optimization model with dynamic voltage balance, energy storage load balance and minimum energy storage dynamic degradation cost as optimization objectives; solving the multi-objective optimization model based on a Bayesian optimization and MPC hybrid algorithm to obtain optimal control instructions; constructing a photovoltaic storage hierarchical collaborative control strategy with photovoltaic priority and energy storage supplement based on the optimal control instructions to complete voltage and load adjustment within the cluster; and for a photovoltaic storage adjustment resource insufficient scene, realizing reactive power and voltage collaborative supplement control of the power distribution network based on AVC regulation of a superior transformer substation. The application can dynamically balance the conflicting objectives of voltage recovery accuracy, energy storage load balance and operation economy.

Description

Technical Field

[0001] This application relates to the field of power system automation, and in particular to a method, system, equipment, medium and product for voltage control of power distribution networks. Background Technology

[0002] With the rapid development of new energy sources, distributed photovoltaic (PV) power, with its advantages of being clean and sustainable, has seen a continuous increase in its integration rate into distribution networks, becoming an important component of energy supply for these networks. However, PV power output exhibits significant volatility and intermittency, and the distribution network has numerous nodes and complex electrical coupling relationships. High-penetration PV integration leads to increased voltage fluctuations and more frequent voltage exceedances, severely impacting power supply quality and the safe and stable operation of the system. Regarding optimization algorithms, existing distribution network voltage control methods often employ a single optimization algorithm to handle multi-objective control problems. A single MPC algorithm struggles to dynamically balance the conflicting objectives of voltage recovery accuracy, energy storage load balancing, and operational economy, failing to adapt to dynamic distribution network operating scenarios. Traditional heuristic algorithms have slow convergence speeds, are prone to getting trapped in local optima, and struggle to output real-time optimal control commands. Summary of the Invention

[0003] The purpose of this application is to provide a distribution network voltage control method, system, equipment, medium and product that can solve the problem that related technologies are difficult to dynamically balance the conflicting goals of voltage recovery accuracy, energy storage load balancing and operation economy, and are prone to getting trapped in local optima and unable to output optimal commands in real time.

[0004] To achieve the above objectives, this application provides the following solution.

[0005] Firstly, this application provides a power distribution network voltage control method, including: Obtain real-time operation data of the power distribution network; The distribution network nodes are divided into real-time clusters based on real-time operation data of the distribution network. A multi-objective optimization model is constructed with the optimization objectives of dynamic voltage balance, energy storage load balancing, and minimizing the dynamic degradation cost of energy storage. The optimal control command is obtained by solving the multi-objective optimization model based on a hybrid algorithm of Bayesian optimization and MPC. Based on the optimal control command, a photovoltaic-storage hierarchical collaborative control strategy with photovoltaic priority and energy storage supplementation is constructed to complete the voltage and load regulation within the cluster; For scenarios where photovoltaic and energy storage regulation resources are insufficient during voltage and load regulation within the cluster, reactive power and voltage collaborative supplementary control of the distribution network is achieved based on the AVC control of the upper-level substation, thus completing the closed-loop stable regulation of the distribution network voltage. The scenario where photovoltaic and energy storage regulation resources are insufficient is when the reactive power output of photovoltaics and the charging and discharging power of energy storage have reached the limit and the node voltage is still exceeding the limit.

[0006] Secondly, this application provides a power distribution network voltage control system, comprising: The data acquisition module is used to collect real-time operating data of the power distribution network; The clustering module is used to perform real-time clustering of distribution network nodes based on real-time operation data of the distribution network. The hybrid optimization module is used to construct a multi-objective optimization model with the optimization objectives of dynamic voltage balance, energy storage load balancing, and minimizing the dynamic degradation cost of energy storage. The multi-objective optimization model is solved based on a hybrid algorithm of Bayesian optimization and MPC to obtain the optimal control command. The photovoltaic-storage coordinated control module is used to construct a photovoltaic-storage hierarchical coordinated control strategy based on the optimal control command, prioritizing photovoltaics and supplementing energy storage, to complete voltage and load regulation within the cluster; The upper and lower level collaborative control module is used to address scenarios where photovoltaic and energy storage regulation resources are insufficient. Based on the AVC regulation of the upper substation, it realizes the collaborative supplementary control of reactive power and voltage in the distribution network, and completes the closed-loop stable regulation of the distribution network voltage.

[0007] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the computer program to implement the steps of the above-described power distribution network voltage control method.

[0008] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described power distribution network voltage control method.

[0009] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described power distribution network voltage control method.

[0010] According to the specific embodiments provided in this application, this application has the following technical effects: This application provides a distribution network voltage control method, system, equipment, medium, and product. By solving a multi-objective optimization model using a hybrid algorithm of Bayesian optimization and MPC, the optimal control command is obtained. This solves the problem that related technologies struggle to dynamically balance conflicting objectives such as voltage recovery accuracy, energy storage load balancing, and operational economy, are prone to getting trapped in local optima, and are unable to output the optimal command in real time. At the same time, a collaborative linkage mechanism is established between the distribution network's photovoltaic-storage cluster and the upstream substation's AVC, thereby ensuring the multi-objective balance control effect in conventional scenarios and solving the voltage stability problem in extreme scenarios with insufficient regulation resources, thus strengthening the distribution network's ability to accept high-penetration renewable energy sources. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly described below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort: Figure 1 This is an application environment diagram of a power distribution network voltage control method according to an embodiment of this application; Figure 2 A schematic flowchart illustrating a power distribution network voltage control method according to an embodiment of this application; Figure 3 This is a schematic diagram illustrating the specific process of a power distribution network voltage control method according to an embodiment of this application; Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0013] As a core technology to ensure the distribution network's ability to accept new energy sources, precise voltage control must simultaneously meet multiple requirements, including regulation response speed, voltage control accuracy, resource utilization efficiency, and operational economy.

[0014] Existing voltage control methods for distribution networks still have many technical limitations and are difficult to adapt to scenarios with high proportions of photovoltaic (PV) and energy storage integration. In the cluster partitioning stage, traditional clustering algorithms often partition based on physical topological distance or single electrical parameters, failing to fully integrate multi-dimensional characteristics such as inter-node electrical coupling strength, adjustable reactive power capacity of PV, and energy storage state of charge. This results in a mismatch between the partitioning results and the actual operating characteristics of the distribution network and the distribution of PV and energy storage resources, leading to a lack of targeted control commands and low regulation efficiency. Furthermore, existing methods generally lack a robust upper-lower level collaborative control mechanism. When the PV and energy storage regulation resources on the distribution network reach their limits but still cannot resolve voltage exceedances, they cannot quickly supplement regulation, making it difficult to guarantee voltage stability in extreme scenarios and restricting the distribution network's ability to accommodate high proportions of new energy. Regarding regulation strategies, some methods rely solely on PV reactive power regulation or activate energy storage regulation alone. The reactive power regulation range of PV inverters is limited, making it difficult to cope with large voltage deviations. When energy storage participates in regulation alone, problems such as load imbalance within and between clusters, and accelerated equipment degradation due to frequent charging and discharging can easily occur, without fully considering the dynamic degradation costs of energy storage.

[0015] Therefore, there is an urgent need for a cluster partitioning method that can accurately match the operating characteristics of the distribution network with the distribution of photovoltaic and energy storage resources, construct a hierarchical collaborative regulation strategy for photovoltaic and energy storage, combine an efficient hybrid optimization algorithm to achieve multi-objective dynamic balance, and establish a collaborative regulation mechanism between upper and lower levels. This will address the shortcomings of existing methods in terms of control accuracy, resource utilization, and scenario adaptability, and provide technical support for the safe, stable, economical, and efficient operation of the distribution network under high-penetration photovoltaic and energy storage access.

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

[0017] The power distribution network voltage control method provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be set up independently, integrated into server 104, or placed in the cloud or on other servers. Terminal 102 can send real-time distribution network operation data to server 104. After receiving the real-time distribution network operation data, server 104 performs real-time clustering of distribution network nodes based on the data. With dynamic voltage balance, energy storage load balancing, and minimizing the dynamic degradation cost of energy storage as optimization objectives, a multi-objective optimization model is constructed. The model is solved using a hybrid algorithm of Bayesian optimization and MPC to obtain the optimal control command. Based on the optimal control command, a photovoltaic-priority, energy storage-supplemented hierarchical collaborative control strategy is constructed to complete voltage and load regulation within the cluster. For scenarios with insufficient photovoltaic and energy storage regulation resources, reactive power and voltage collaborative supplementary control of the distribution network is achieved based on the AVC regulation of the upper-level substation, completing the closed-loop stable regulation of the distribution network voltage. Server 104 can feed back the optimal control commands obtained based on the real-time operation data of the distribution network to terminal 102. Furthermore, in some embodiments, the distribution network voltage control method can also be implemented independently by server 104 or terminal 102. For example, terminal 102 can directly perform distribution network voltage control based on the real-time operation data of the distribution network, or server 104 can obtain the real-time operation data of the distribution network from the data storage system and perform distribution network voltage control based on the real-time operation data of the distribution network.

[0018] The terminal 102 can be, but is not limited to, various desktop computers and laptops. The server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers, or it can be a cloud server.

[0019] In one exemplary embodiment, such as Figure 2As shown, a distribution network voltage control method is provided. This method is executed by computer equipment, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is applied to... Figure 1 Taking server 104 as an example, the explanation includes the following steps 201 to 206.

[0020] Step 201: Obtain real-time operation data of the power distribution network.

[0021] Step 202: Perform real-time clustering of distribution network nodes based on real-time operation data of the distribution network.

[0022] Step 203: With the optimization objectives of dynamic voltage balance, energy storage load balancing and minimizing the dynamic degradation cost of energy storage, a multi-objective optimization model is constructed.

[0023] Step 204: Solve the multi-objective optimization model based on the hybrid algorithm of Bayesian optimization and MPC (Model Predictive Control) to obtain the optimal control command.

[0024] Step 205: Based on the optimal control command, construct a photovoltaic-priority, energy storage-supplemented hierarchical coordinated control strategy to complete voltage and load regulation within the cluster.

[0025] Step 206: For scenarios where photovoltaic and energy storage regulation resources are insufficient during voltage and load regulation within the cluster, reactive power and voltage collaborative supplementary control of the distribution network is achieved based on the AVC (Automatic Voltage Control) regulation of the upper-level substation, thus completing the closed-loop stable regulation of the distribution network voltage. The scenario where photovoltaic and energy storage regulation resources are insufficient is a scenario where the reactive power output of photovoltaics and the charging and discharging power of energy storage have reached the limit and the node voltage is still exceeding the limit.

[0026] By implementing steps 201 to 206 above, a hybrid Bayesian optimization and MPC algorithm is used to solve the multi-objective optimization model, obtaining the optimal control command. This solves the problem that related technologies struggle to dynamically balance conflicting objectives such as voltage recovery accuracy, energy storage load balancing, and operational economy, easily getting trapped in local optima and unable to output the optimal command in real time. Simultaneously, a collaborative linkage mechanism is established between the distribution network's photovoltaic-storage cluster and the upstream substation's AVC (Automatic Ventilation Control). This ensures multi-objective balance control performance in conventional scenarios while addressing voltage stability issues in extreme scenarios with insufficient regulation resources, strengthening the distribution network's ability to accommodate high-penetration renewable energy sources. By improving the reactive power and voltage control accuracy of the distribution network through the hybrid Bayesian optimization and MPC algorithms, and ensuring voltage stability in extreme scenarios with insufficient regulation resources through upstream and downstream collaborative control, this provides technical support for the safe and stable operation of distribution networks with high proportions of renewable energy and energy storage devices.

[0027] At the transformer substation outlet, smart terminals are deployed to collect time-series data on voltage, current, and power. At distributed photovoltaic (PV) access nodes, IoT meters are deployed to collect PV installation location data, rated capacity, reactive power regulation range, real-time output, reactive power output, and power factor data. At user-end meter boxes, LTU (linear threshold unit) sensing units are deployed to collect user-side voltage and current time-series data. At energy storage access nodes, energy storage SOC (State of Charge) sensing modules are deployed to collect energy storage installation location data, rated capacity, charge / discharge power range, real-time SOC, and charge / discharge power data. The specific process of distribution network voltage control is as follows: Figure 3 As shown.

[0028] In step 201 above, the real-time operation data of the distribution network specifically includes: distribution area outlet voltage, current, power time-series data, photovoltaic installation location data and photovoltaic rated capacity, reactive power regulation range, real-time output, reactive power output and power factor data, user-side voltage, current time-series data, energy storage installation location data and energy storage rated capacity, charging and discharging power range, real-time SOC, and charging and discharging power data.

[0029] Distribution network nodes are divided into three categories: ordinary distribution network nodes that only bear loads, distributed photovoltaic access nodes with inverter regulation capabilities, and energy storage access nodes with charging, discharging, and SOC characteristics.

[0030] In step 202 above, the dynamic clustering algorithm based on Graph Convolutional Network (GCN) divides the distribution network nodes into three categories for real-time cluster partitioning. This real-time cluster partitioning of distribution network nodes based on real-time operation data specifically includes: inputting the real-time operation data into the trained GCN model and outputting the cluster to which each node belongs; extracting the basic coupling strength value based on the adjacency matrix, combining it with the intermediate features output from the graph convolutional layers in the GCN model, and obtaining the coupling coefficient between distribution network nodes through feature similarity correction to perform cluster partitioning of the distribution network nodes.

[0031] After inputting real-time operation data of the distribution network into the trained GCN model, the GCN model needs to be trained. The specific process includes the following steps 301 to 305.

[0032] Step 301: Construct an undirected graph model of the distribution network and define the undirected graph of the distribution network. ,in It is a set of nodes that includes distribution network nodes, photovoltaic access nodes, and energy storage access nodes. Let be the set of edges corresponding to the distribution network lines, where Represents a node and Directly connected, otherwise =0 indicates a node and Not connected.

[0033] Step 302: Extract multi-dimensional feature vectors of nodes, extracting the 8-dimensional feature vector for each node. Specifically, this includes node voltages. Active power reactive power Photovoltaic power output Adjustable reactive power capacity of photovoltaics Energy storage state of charge Reactive voltage sensitivity Active voltage sensitivity ; It is the set of real numbers.

[0034] Step 303: Calculate the adjacency matrix based on the electrical distance between nodes Define nodes and The adjacency weight between them is The expression is as follows: (1); in, For nodes and Electrical distance between them This refers to the maximum electrical distance in the power distribution network.

[0035] Step 304: Design the GCN network hierarchical structure, input layer receive node feature matrix (Including the 8-dimensional eigenvectors of all nodes) and the adjacency matrix The convolutional layer in the figure uses The graph convolution kernel approximated by the Chebyshev polynomial is expressed as follows: (2); in, The intermediate features output by the graph convolutional layer; For the normalized Laplace matrix; It is a Laplace matrix; for The largest eigenvalue; It is the identity matrix; for Step Polynomial; These are the kernel weights; for Activation function, polynomial order .

[0036] Step 305: Construct the training dataset for the GCN clustering model, including node feature matrices for different operating scenarios of the distribution network. and adjacency matrix Multiple sets of samples are used, each with a label including the cluster affiliation label of a node in the distribution network. Training parameters are set, and the cross-entropy loss function is used. The trained GCN clustering model (i.e., the pre-trained GCN model) is expressed as follows: (3); in, The number of nodes in the distribution network; Represents a node The cluster to which it belongs is c. Represents a node The cluster to which it belongs is not c; Number of clusters; For nodes The probability that it is predicted to belong to cluster c.

[0037] The real-time collected distribution network operation data is input into the trained GCN model, which outputs the affiliation probability of each node. The cluster to which it belongs is The output nodes of the trained GCN model The probability of being predicted as a node for each cluster is used to determine the cluster with the highest median probability. The cluster to which it belongs.

[0038] Based on adjacency matrix Extract the basic values ​​of coupling strength and combine them with the intermediate features output by the graph convolutional layer. By correcting for feature similarity, the coupling coefficient between distribution network nodes is obtained. Perform cluster partitioning of distribution network nodes, nodes and Coupling coefficient between The expression is as follows: (4); in, , They are nodes and In the output feature vector of the graph convolutional layer, a threshold is set. , The coupling levels are divided into strong, medium and weak to determine the coupling levels between any two nodes in the distribution network. In order to combine the optimal control commands, a photovoltaic-priority and energy storage-supplemented hierarchical collaborative control strategy is constructed.

[0039] The coupling degree between distribution network nodes specifically refers to the degree of electrical interconnection between nodes caused by line connections. It is reflected in the impact of voltage and power changes at one node on other nodes. The coupling degree between distribution network nodes is measured by the coupling coefficient. Quantification, that is The larger the value, the tighter the electrical coupling between nodes.

[0040] Determining the coupling degree ensures strong electrical connections between nodes within the same cluster, enabling rapid transmission of control effects during regulation. Prioritizing the control of nodes with strong coupling levels allows for global voltage recovery with minimal adjustment, reducing ineffective control. Nodes with medium coupling levels serve as auxiliary control targets. After identifying nodes with weak coupling levels, these nodes do not need to follow the control actions of other clusters. They only participate in control when voltage exceeds limits occurs in their own node or cluster, avoiding voltage oscillations caused by cross-cluster control.

[0041] Based on the cluster division results, the distribution characteristics of photovoltaic and energy storage resources are screened. Specifically, the distribution characteristics of photovoltaic and energy storage resources are as follows: Spatial distribution characteristics: distribution network node numbers including photovoltaic and energy storage access, and the number of photovoltaic and energy storage nodes in each cluster; Capacity distribution characteristics: total adjustable reactive power capacity of photovoltaic and total energy storage capacity in each cluster; Dynamic distribution characteristics: real-time total output of photovoltaic and real-time average SOC of energy storage in each cluster.

[0042] The multi-objective optimization model for step 203 above is expressed as follows: (5); (6); (7); (8); , (9); in, For multi-objective optimization models; These are adaptive weighting coefficients; The objective function is the voltage deviation. For the first The voltage of each node; The number of nodes in the distribution network; Rated voltage; The objective function for energy storage load balancing is... The set of energy storage access nodes participating in regulation; This is the number of any energy storage node in the set of energy storage access nodes; This represents the real-time load rate of the energy storage node. The average load factor of all energy storage nodes participating in the regulation; The objective function for the dynamic degradation cost of energy storage; For the first The dynamic degradation cost of an energy storage device during the current control cycle; Number the energy storage device; The number of energy storage devices participating in regulation; For the first The state of charge of an energy storage device; and The first One energy storage device in and The state of charge at any given moment; For the first The charging and discharging power of an energy storage device, with positive and negative signs used to distinguish between charging and discharging, and the absolute value of the power in the formula represents the charging and discharging power amplitude. To control cycle duration; For the first An energy storage device in a state of charge The unit charge-discharge degradation cost function is as follows; and The first An energy storage device in a state of charge and The unit charge / discharge degradation cost; For the first The cost of purchasing or replacing an energy storage device; For the first The rated capacity of each energy storage device; For the first The charging and discharging efficiency of an energy storage device; parameters , These are parameters for the energy storage lifetime degradation model, used to characterize the impact of changes in the energy storage state of charge on degradation costs.

[0043] In step 204 above, the multi-objective optimization model is solved based on the hybrid algorithm of Bayesian optimization and MPC to obtain the optimal control command, specifically including the following steps 401 to 402.

[0044] Step 401: Using the adaptive weight coefficients as optimization variables, construct a surrogate model, perform Bayesian optimization using a Gaussian process model, and iteratively optimize to obtain the optimal adaptive weights for each objective function; the surrogate model is a Gaussian process probability approximation function that takes the adaptive weight vector as input and outputs the multi-objective optimization model function value after MPC rolling optimization.

[0045] Define the optimization variables as weight vectors Build a proxy model, , , The weights corresponding to the voltage deviation target, energy storage load balancing target, and energy storage degradation cost target are respectively used, and a Gaussian process model is employed for Bayesian iterative optimization. As a convergence condition, the optimal adaptive weights are output through iterative optimization. The kernel function expression is as follows: (10); in, is the kernel function of the surrogate model, used to measure the correlation between two candidate weight vectors; exp() is the natural exponential function; The signal variance parameter of the kernel function; and For any two candidate weight vectors during the Bayesian optimization iteration process; and Weight vectors and In the Weight components in each target dimension; The length scale parameter of the kernel function is used to characterize the impact of changes in the weight vector on the correlation of the objective function value; These are the adaptive weights obtained in the (k+1)th and kth iterations, respectively.

[0046] Step 402: Based on the optimal adaptive weight, the multi-objective function is transformed into a weighted single-objective function, and the interior point method is used to solve the MPC quadratic programming problem. With photovoltaic reactive power output and energy storage charging and discharging power as control variables, the photovoltaic-storage control command sequence in the control time domain is updated by rolling optimization, and the optimal control command is output.

[0047] MPC optimization is performed to transform the multi-objective function into a weighted single-objective function. The interior-point method is used to solve the quadratic programming problem, and the optimal command sequence for optical storage control is output. The weighted single-objective function expression is as follows: (11); in, The optimal adaptive weights obtained in step 401; the reactive power output of the photovoltaic access node and the charging and discharging power of the energy storage access node are the control variables in the MPC optimization process, influenced by power flow prediction and energy storage state prediction. , , .

[0048] The Bayesian iterative optimization and MPC optimization solutions are specifically implemented by adaptively adjusting weights through Bayesian optimization under different operating scenarios. The goal is to achieve a comprehensive optimal balance with the objectives of minimizing voltage deviation, minimizing energy storage load rate differences, and minimizing energy storage degradation costs. When solving the problem through MPC optimization, the reactive power output of each photovoltaic access node and the charging and discharging power of each energy storage access node are used as control variables in the predictive control time domain. The node voltage change, energy storage load rate change, and energy storage SOC change under the action of these control variables are substituted into the voltage deviation target, energy storage load balancing target, and energy storage degradation cost target, respectively, for rolling optimization evaluation. In this way, the multi-objective requirements based on Bayesian optimization are transformed into specific operation instructions that photovoltaic and energy storage devices can directly execute, i.e., optimal control instructions. The optimal control instructions include the reactive power output sequence of each photovoltaic access node and the charging and discharging power sequence of each energy storage access node. The length of the sequence is the control time domain of MPC and is updated in real time according to the rolling optimization logic.

[0049] The photovoltaic-storage hierarchical coordinated control strategy in step 205 above specifically includes: 1) Prioritize photovoltaic reactive power regulation, control the output reactive power of the photovoltaic inverter according to the optimal control command, and complete the voltage regulation if the photovoltaic regulation constraints are met and the node voltage returns to the qualified range.

[0050] Based on the reactive power output sequence of the photovoltaic access node, control the output of the photovoltaic inverter. If photovoltaic reactive power regulation is implemented and photovoltaic regulation constraints are met... And the voltage of the original voltage limit-exceeding node satisfies Then the adjustment is complete; among them, These represent the minimum and maximum power output of the photovoltaic inverter, respectively. The voltages of the nodes that exceeded the voltage limit are respectively. The minimum and maximum values.

[0051] 2) If the photovoltaic reactive power output reaches the regulation limit but the voltage exceeds the limit, the energy storage supplementary regulation is activated. The corresponding energy storage charging and discharging power calculation formula is matched according to the voltage exceeding the upper and lower limits, and the energy storage regulation power is output.

[0052] If the photovoltaic inverter output If the voltage exceeds the photovoltaic regulation constraint, the system will switch to energy storage for supplementary regulation. This applies to scenarios where the voltage exceeds the upper limit, specifically the energy storage charging power. The expression is as follows: (12); in, This indicates taking the minimum value; This represents the maximum value of the energy storage charging and discharging power. Initial energy storage charging power; The voltage at the node; The energy storage power-voltage regulation sensitivity coefficient is the regulation coefficient between voltage deviation and energy storage charging and discharging power, used to characterize the amount of energy storage power regulation corresponding to a unit voltage deviation.

[0053] Energy storage discharge power in scenarios where the voltage is below the lower limit The expression is as follows: (13); in, This indicates taking the maximum value; This represents the minimum value of the energy storage charging and discharging power.

[0054] 3) Energy storage is adjusted according to the optimal control command. If the voltage does not meet the standard after adjustment, it is adjusted again by the charging and discharging power formula of the matching scenario until the node voltage meets the operating requirements.

[0055] After the energy storage is adjusted according to the charging and discharging power sequence of the energy storage access node, if the voltage of the original voltage-over-limit node meets the requirements... The adjustment is then complete. If the requirements are still not met, the energy storage charging power will be adjusted according to the voltage limit mentioned above. Or the energy storage discharge power when the voltage is below the lower limit Supplementary adjustments will be made.

[0056] Step 206 above is the voltage control process under the scenario of insufficient photovoltaic and energy storage regulation resources. For the scenario of insufficient photovoltaic and energy storage regulation resources, reactive power and voltage collaborative supplementary control of the distribution network is realized based on the AVC regulation of the upper substation to complete the closed-loop stable regulation of the distribution network voltage. Specifically, it includes the following steps 501 to 505.

[0057] Step 501: When the reactive power output of photovoltaic power and the charging and discharging power of energy storage have reached the limit and the node voltage is still exceeding the limit, it is determined to be a scenario where photovoltaic and energy storage regulation resources are insufficient.

[0058] The expression for determining insufficient photovoltaic and energy storage regulation resources is as follows: (14); in, For nodes The reactive power output of photovoltaics.

[0059] Step 502, based on rated voltage and voltage tolerance Calculate the voltage coordination limits of the AVC system of the upstream substation. The expression is as follows: (15); in, For the first The voltage of each node.

[0060] Step 503: The voltage coordination limit, insufficient regulation resource alarm signal and real-time operation data of the distribution network of the upper substation AVC system are encrypted and transmitted to the upper substation AVC system through distribution communication.

[0061] Step 504: The upstream substation AVC system formulates a control strategy based on the voltage coordination limit, and supplements the reactive power of the system by switching on and off parallel capacitor banks, parallel reactor banks or adjusting the tap changer of the on-load tap-changing transformer, thereby adjusting the bus voltage to the voltage coordination limit.

[0062] After receiving the signal, the AVC system of the upstream substation formulates a control strategy based on the voltage coordination limit. When the voltage exceeds the lower limit, it switches parallel capacitor banks or adjusts the tap changer of the on-load tap-changing transformer to supplement the reactive power of the system and raise the bus voltage to the required level. In scenarios where the voltage exceeds the upper limit, switching parallel reactors or adjusting the tap changer of the on-load tap-changing transformer can absorb excess reactive power in the system and reduce the bus voltage to [a lower limit]. .

[0063] Step 505: The distribution network side feeds back node voltage data to the upper-level AVC system once per cycle. When the voltage of all nodes in the distribution network recovers to the acceptable range and remains so for a set time period, a control termination signal is sent. The set time period can be 3 sampling cycles. Furthermore, after three consecutive sampling cycles, a control termination signal is sent, and the AVC system of the upper-level substation stops reactive power regulation.

[0064] This application has the following beneficial effects: First, this application considers that traditional clustering algorithms rely only on physical topology or a single electrical parameter, which makes it difficult to match the dynamic operating characteristics of the distribution network with the distribution of photovoltaic and energy storage resources, resulting in a lack of targeted voltage regulation. Therefore, a dynamic clustering algorithm based on graph convolutional networks is adopted to divide the clusters and construct a photovoltaic-energy storage hierarchical control strategy, thereby achieving precise matching of photovoltaic and energy storage resources with voltage regulation needs, giving full play to the advantages of photovoltaic rapid reactive power support and energy storage flexible regulation, and improving the targetedness and flexibility of voltage control.

[0065] Second, considering that a single optimization algorithm is difficult to balance multiple objectives such as voltage control accuracy and operational economy, and that conventional control lacks a mechanism to supplement regulation resources in extreme scenarios, this application constructs a multi-objective optimization model and solves the optimal control command based on a Bayesian optimization + MPC hybrid algorithm. At the same time, a collaborative linkage mechanism between the distribution network photovoltaic-storage cluster and the upper-level substation AVC is established. This not only ensures the multi-objective balanced control effect in conventional scenarios, but also solves the voltage stability problem in extreme scenarios with insufficient regulation resources, thereby strengthening the distribution network's ability to accept high-penetration new energy sources.

[0066] This application also provides an application scenario in which the above-described distribution network voltage control method is applied. Specifically, the distribution network voltage control method provided in this embodiment can be applied in a power system automation control scenario. The power system automation control scenario includes a content production stage, a distribution network voltage control link, and a power system automation control stage. Real-time operation data of the distribution network enters the distribution network voltage control link from the content production stage, obtains corresponding content features through human-machine collaboration, and then enters the downstream power system automation control stage. The distribution network voltage control method provided in this embodiment belongs to the distribution network voltage control link. Specifically, in the distribution network voltage control link process targeting real-time operation data of the distribution network, the distribution network nodes can be divided into real-time clusters based on the real-time operation data of the distribution network. With the optimization objectives of dynamic voltage balance, energy storage load balancing, and minimizing the dynamic degradation cost of energy storage, a multi-objective optimization model is constructed. The multi-objective optimization model is solved based on a hybrid algorithm of Bayesian optimization and MPC to obtain the optimal control command. Based on the optimal control command, a photovoltaic-energy storage hierarchical collaborative control strategy is constructed to complete the voltage and load regulation within the cluster. For scenarios where photovoltaic and energy storage regulation resources are insufficient, reactive power and voltage collaborative supplementary control of the distribution network is realized based on the AVC regulation of the upper-level substation, thus completing the closed-loop stable regulation of the distribution network voltage.

[0067] Based on the same inventive concept, this application also provides a distribution network voltage control system for implementing the aforementioned distribution network voltage control method. The solution provided by this system is similar to the implementation scheme described in the above method; therefore, the specific limitations in one or more embodiments of the distribution network voltage control system provided below can be found in the limitations of the distribution network voltage control method described above, and will not be repeated here.

[0068] In one exemplary embodiment, a power distribution network voltage control system is provided, including a data acquisition module, a clustering module, a hybrid optimization module, a photovoltaic-storage collaborative control module, and an upper-lower level collaborative control module.

[0069] The data acquisition module is used to collect real-time operating data of the distribution network. Specifically, it collects the voltage, power, photovoltaic output, energy storage SOC and line parameters of the distribution network nodes in real time and transmits them to other modules in encrypted form.

[0070] The data acquisition module includes a smart terminal for the distribution area, IoT meters, LTU sensing units, and an energy storage SOC sensing module. The specific installation locations are as follows: the smart terminal for the distribution area is deployed at the distribution area outlet, such as the transformer JP cabinet; the IoT meters are deployed at the distributed photovoltaic access nodes, such as the inverter outlet; the LTU sensing units are deployed at the user end meter boxes; and the energy storage SOC sensing modules are deployed at the energy storage access nodes, such as the converter outlet.

[0071] The clustering module is used to perform real-time clustering of distribution network nodes based on real-time operation data of the distribution network.

[0072] The hybrid optimization module is used to construct a multi-objective optimization model with the optimization objectives of dynamic voltage balance, energy storage load balancing, and minimizing the dynamic degradation cost of energy storage. The multi-objective optimization model is solved based on a hybrid algorithm of Bayesian optimization and MPC to obtain the optimal control command.

[0073] The photovoltaic-storage coordinated control module is used to construct a photovoltaic-storage hierarchical coordinated control strategy based on the optimal control command, and to complete the voltage and load regulation within the cluster.

[0074] The upper and lower level collaborative control module is used to address scenarios where photovoltaic and energy storage regulation resources are insufficient. Based on the AVC regulation of the upper substation, it realizes the collaborative supplementary control of reactive power and voltage in the distribution network, and completes the closed-loop stable regulation of the distribution network voltage.

[0075] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 4 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores power distribution network voltage control data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a power distribution network voltage control method.

[0076] Figure 4 The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0077] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0078] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0079] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of related data must comply with relevant regulations and be authorized by the owner of the corresponding device.

[0080] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0081] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0082] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0083] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A voltage control method for a power distribution network, characterized in that, The power distribution network voltage control method includes: Obtain real-time operation data of the power distribution network; The distribution network nodes are divided into real-time clusters based on real-time operation data of the distribution network. A multi-objective optimization model is constructed with the optimization objectives of dynamic voltage balance, energy storage load balancing, and minimizing the dynamic degradation cost of energy storage. The optimal control command is obtained by solving the multi-objective optimization model based on a hybrid algorithm of Bayesian optimization and MPC. Based on the optimal control command, a photovoltaic-storage hierarchical collaborative control strategy with photovoltaic priority and energy storage supplementation is constructed to complete the voltage and load regulation within the cluster; For scenarios where photovoltaic and energy storage regulation resources are insufficient during voltage and load regulation within the cluster, reactive power and voltage collaborative supplementary control of the distribution network is achieved based on the AVC control of the upper-level substation, thus completing the closed-loop stable regulation of the distribution network voltage. The scenario where photovoltaic and energy storage regulation resources are insufficient is when the reactive power output of photovoltaics and the charging and discharging power of energy storage have reached the limit and the node voltage is still exceeding the limit.

2. The power distribution network voltage control method according to claim 1, characterized in that, The distribution network nodes are divided into real-time clusters based on real-time operation data, specifically including: Input the real-time operation data of the distribution network into the trained GCN model, and output the cluster to which each node belongs; Based on the adjacency matrix, the basic values ​​of coupling strength are extracted. Combined with the intermediate features output by the graph convolutional layer in the GCN model, the coupling coefficients between distribution network nodes are obtained through feature similarity correction, and the distribution network nodes are clustered.

3. The power distribution network voltage control method according to claim 1, characterized in that, The multi-objective optimization model is represented as follows: ； ； ； ； ， ； in, For multi-objective optimization models; These are adaptive weighting coefficients; The objective function is the voltage deviation. For the first The voltage of each node; The number of nodes in the distribution network; Rated voltage; The objective function for energy storage load balancing is... The set of energy storage access nodes participating in regulation; This is the number of any energy storage node in the set of energy storage access nodes; This represents the real-time load rate of the energy storage node. The average load factor of all energy storage nodes participating in the regulation; The objective function for the dynamic degradation cost of energy storage; For the first The dynamic degradation cost of an energy storage device during the current control cycle; Number the energy storage device; The number of energy storage devices participating in the regulation; For the first The state of charge of an energy storage device; and The first One energy storage device in and The state of charge at any given moment; For the first The charging and discharging power of an energy storage device; To control cycle duration; For the first An energy storage device in a state of charge The unit charge-discharge degradation cost function is as follows; and The first An energy storage device in a state of charge and The unit charge / discharge degradation cost; For the first The cost of purchasing or replacing an energy storage device; For the first The rated capacity of each energy storage device; For the first The charging and discharging efficiency of an energy storage device; parameters , These are parameters for the energy storage lifetime degradation model, used to characterize the impact of changes in the energy storage state of charge on degradation costs.

4. The power distribution network voltage control method according to claim 3, characterized in that, The multi-objective optimization model is solved using a hybrid algorithm combining Bayesian optimization and MPC to obtain the optimal control command, specifically including: Using adaptive weight coefficients as optimization variables, a surrogate model is constructed, and Bayesian optimization is performed using a Gaussian process model. The optimal adaptive weights for each objective function are obtained through iterative optimization. The surrogate model is a Gaussian process probability approximation function that takes the adaptive weight vector as input and outputs the multi-objective optimization model function value after MPC rolling optimization. Based on the optimal adaptive weights, the multi-objective function is transformed into a weighted single-objective function, and the interior point method is used to solve the MPC quadratic programming problem. Using photovoltaic reactive power output and energy storage charging and discharging power as control variables, the photovoltaic-energy storage control command sequence in the control time domain is updated by rolling optimization to output the optimal control command.

5. The power distribution network voltage control method according to claim 1, characterized in that, The photovoltaic-storage hierarchical coordinated control strategy specifically includes: Prioritize photovoltaic reactive power regulation, control the photovoltaic inverter to output reactive power according to the optimal control command, and complete the voltage regulation if the photovoltaic regulation constraints are met and the node voltage returns to the qualified range. If the reactive power output of photovoltaics reaches the regulation limit but the voltage exceeds the limit, the energy storage supplementary regulation is activated. The corresponding energy storage charging and discharging power calculation formula is matched according to the voltage exceeding the upper and lower limits, and the energy storage regulation power is output. Energy storage is adjusted according to the optimal control command first. If the voltage does not meet the standard after adjustment, it is adjusted again by the charging and discharging power formula matching the scenario until the node voltage meets the operating requirements.

6. The power distribution network voltage control method according to claim 1, characterized in that, To address scenarios where photovoltaic and energy storage regulation resources are insufficient, reactive power and voltage coordinated supplementary control of the distribution network is achieved based on the AVC control of the upstream substation, thus completing the closed-loop stability control of the distribution network voltage. Specifically, this includes: When the reactive power output of photovoltaics and the charging and discharging power of energy storage have reached the limit and the node voltage is still exceeding the limit, it is judged as a scenario where photovoltaic and energy storage regulation resources are insufficient. Calculate the voltage coordination limit of the AVC system of the upstream substation based on the rated voltage and the allowable voltage deviation; The voltage coordination limit, insufficient regulation resource alarm signal and real-time operation data of the distribution network are encrypted and transmitted to the AVC system of the upper substation. The upstream substation AVC system formulates control strategies based on voltage coordination limits, supplementing the system's reactive power by switching on and off parallel capacitor banks, parallel reactor banks, or adjusting the tap changers of on-load tap-changing transformers, thereby adjusting the bus voltage to the voltage coordination limits. The distribution network side feeds back node voltage data to the upper-level AVC system once per cycle. When the voltage of all nodes in the distribution network recovers to the qualified range and continues for a set period of time, a control termination signal is sent.

7. A power distribution network voltage control system, used to execute the power distribution network voltage control method according to any one of claims 1-6, characterized in that, The power distribution network voltage control system includes: The data acquisition module is used to collect real-time operating data of the power distribution network; The clustering module is used to perform real-time clustering of distribution network nodes based on real-time operation data of the distribution network. The hybrid optimization module is used to construct a multi-objective optimization model with the optimization objectives of dynamic voltage balance, energy storage load balancing, and minimizing the dynamic degradation cost of energy storage. The multi-objective optimization model is solved based on a hybrid algorithm of Bayesian optimization and MPC to obtain the optimal control command. The photovoltaic-storage coordinated control module is used to construct a photovoltaic-storage hierarchical coordinated control strategy based on the optimal control command, prioritizing photovoltaics and supplementing energy storage, to complete voltage and load regulation within the cluster; The upper and lower level collaborative control module is used to address scenarios where photovoltaic and energy storage regulation resources are insufficient. Based on the AVC regulation of the upper substation, it realizes the collaborative supplementary control of reactive power and voltage in the distribution network, and completes the closed-loop stable regulation of the distribution network voltage.

8. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that the processor executes the computer program to implement the power distribution network voltage control method according to any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the power distribution network voltage control method according to any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the power distribution network voltage control method according to any one of claims 1-6.

Images