Voltage coordination control method and system based on distributed optical storage

Abstract

The application discloses a voltage coordination control method and system based on distributed optical storage, relates to the technical field of voltage coordination control, and comprises the following steps: acquiring first data of nodes, and constructing local adjustment trend vectors of the nodes based on time windows; based on the adjustment trend vectors, adjustment role division is carried out; based on the division result, a voltage mutual influence graph of local node groups is constructed, behavior conflicts in adjustment are evaluated, and a conflict relationship is obtained; based on the adjustment role and the conflict relationship, a strategy optimization model is constructed, and optimal adjustment strategy combinations of the nodes are solved; each optical storage node executes differentiated voltage control actions according to the assigned adjustment role and specific strategy, and automatically adjusts the response mode according to the behavior change of adjacent nodes. Through the construction of the voltage mutual influence graph and the conflict path identification mechanism, the application can dynamically adjust the response mode under the conditions of communication delay and external disturbance, and maintains the continuity and coordination of voltage control.

Description

Technical Field

[0001] This invention relates to the field of voltage coordination control technology, and more specifically, to a voltage coordination control method and system based on distributed optical energy storage. Background Technology

[0002] With the large-scale integration of distributed photovoltaic (PV) and energy storage systems, the operating characteristics of power distribution networks are undergoing profound changes. Especially in medium- and low-voltage power grids, traditional unidirectional power flow and centralized control systems are struggling to cope with the new power source configurations characterized by multi-source heterogeneity and high volatility. Against this backdrop, "PV-energy storage systems," which combine distributed PV and energy storage, are gradually becoming important participants in voltage regulation, possessing advantages such as high flexibility, fast response speed, and local control, and have broad prospects for engineering applications.

[0003] However, existing voltage control strategies based on distributed photovoltaic (PV) and energy storage systems (ESS) mostly focus on single-node optimal control, passive response based on voltage thresholds, or partitioned coordination through centralized scheduling platforms, without fully considering the coupling relationship between the regulation behaviors of nodes. In scenarios where multiple nodes simultaneously carry out voltage regulation, the lack of differentiated strategy identification mechanisms and conflict prevention measures can easily lead to a "reverse synergy effect"—that is, multiple nodes taking similar regulation measures in the absence of coordination, resulting in voltage over-regulation, oscillation, or regulation failure. In addition, when responding to voltage control tasks, PV-Energy storage systems also need to consider objectives such as SOC protection, frequency support, and energy management, further exacerbating control strategy conflicts and execution complexity.

[0004] For example, the invention patent with publication number CN119518814A describes a method for coordinated control of grid transient voltage under high penetration conditions of distributed photovoltaic and energy storage. This method detects voltage changes at the grid connection point of the distributed photovoltaic power station and, combined with the short-circuit ratio of the grid connection point and the collection point, pre-estimates the reactive power compensation required near the grid connection point at key locations. Then, through coordinated control of photovoltaic and energy storage, and considering different photovoltaic reactive power margin conditions, it develops the reactive power compensation capability of the photovoltaic inverter and achieves smooth processing of its output power, thereby improving the stability of transient voltage at the grid connection point.

[0005] For example, utility model CN216056345U discloses a distributed energy storage coordinated control system based on voltage optimization mode. With energy storage participating in voltage regulation control, it effectively suppresses voltage fluctuations in the power grid, achieving rapid, continuous, and significant voltage adjustments, thus solving the problem of unstable voltage nodes in the power grid. The architecture of the distributed energy storage coordinated control system based on voltage optimization mode can be divided into four layers: local equipment layer, network communication layer, centralized control layer, and application layer. The local equipment layer is used for data acquisition and control of wind power generation units, photovoltaic power generation units, energy storage units, and user loads; the network communication layer is used for real-time transmission of data acquired by the equipment; the centralized control layer is used for coordinated control of the system to maintain the stability of system voltage and frequency; and the application layer is used to display data and provide a visualization platform.

[0006] The above-disclosed technical solutions have at least the following technical problems: when multiple nodes "simultaneously inject reactive power" or reduce active power when the voltage is too high, over-regulation may occur, which may have a reverse effect on neighboring nodes, causing the system to fall into a regulation-anti-regulation-over-regulation cycle.

[0007] To address the above problems, this invention proposes a solution. Summary of the Invention

[0008] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a voltage coordination control method and system based on distributed photovoltaic energy storage. By constructing a node adjustment trend vector, behavioral role recognition, and hierarchical adjustment game model, combined with a voltage mutual influence spectrum and conflict relationship recognition mechanism, and further introducing a differentiable combination strategy optimization and differentiated trajectory deployment mechanism based on heterogeneous graph structure, the present invention addresses the problem in distributed photovoltaic energy storage grid-connected scenarios where multiple nodes exhibit disordered adjustment behavior, overlapping responses, or mutual interference under abnormal voltage conditions, which can easily lead to the system falling into a "adjustment-anti-adjustment-over-adjustment" cycle.

[0009] To achieve the above objectives, the present invention provides the following technical solution:

[0010] The voltage coordination control method based on distributed photovoltaic energy storage includes the following steps: acquiring the first data of each node and constructing a local regulation trend vector for each node based on a time window; identifying the behavioral role of each node in the regulation group based on the regulation trend vector and classifying the regulation roles; constructing a voltage mutual influence map of the local node group based on the classification results, identifying node groups with positive interference and negative cancellation paths, evaluating behavioral conflicts in regulation, and obtaining conflict relationships; constructing a strategy optimization model based on the regulation roles and conflict relationships, and solving for the optimal combination of regulation strategies for each node; each photovoltaic energy storage node executes differentiated voltage control actions according to the assigned regulation role and specific strategy, and automatically adjusts its response mode according to changes in the behavior of neighboring nodes.

[0011] In a preferred embodiment, the step of acquiring the first data of the node and constructing the local adjustment trend vector of each node based on the time window specifically involves: acquiring the first data of the node based on a preset time window length; performing linear regression processing on the first data within the time sliding window to extract the change trend; enhancing the change trend and fusing the voltage phasor difference between the node and its upstream and downstream nodes to construct a local adjustment trend vector characterizing the node's adjustment sensitivity, directionality, and intervention intensity.

[0012] In a preferred embodiment, the step of identifying the behavioral roles of each node in the regulation group based on the regulation trend vector and classifying the regulation roles specifically involves: constructing a topology graph of each distributed optical-storage node based on electrical topology and voltage phasor sensing data, and performing weighted representation to obtain a weighted graph structure; using the local regulation trend vector of each node and a preset response inertia index as the initial node feature vector of the graph neural network; constructing a weighted graph neural network model based on the weighted graph structure and the initial node feature vector to generate a behavioral role embedding vector for each node; constructing a hierarchical regulation game model based on the behavioral role embedding vector to generate the regulation game score for each node; and generating role labels containing regulation priorities based on the regulation game scores of the nodes.

[0013] In a preferred embodiment, the step of constructing a voltage interaction map of a local node group based on the division results specifically involves: aggregating photovoltaic and energy storage nodes that meet preset conditions into a local adjustment node group based on the adjustment role division results; constructing a disturbance conduction channel mapping matrix based on historical voltage monitoring data and node output change data for each pair of nodes within the local node group; and forming a voltage interaction map based on the disturbance conduction channel mapping matrix and the node behavior trend data within the time window.

[0014] In a preferred embodiment, identifying the node group with positive interference and negative cancellation paths specifically involves: generating a voltage response coupling tensor between nodes by calculating the product of the voltage regulation sensitivity matrix and the regulation trend vector between each node in the local node group; constructing a set of directed influence paths based on the coupling tensor; determining that paths with the same edge weights and positive trend correlation are positive interference paths, and paths with opposite edge weights and negative trend correlation are negative cancellation paths.

[0015] In a preferred embodiment, the assessment of behavioral conflicts in the adjustment process to obtain conflict relationships specifically involves: identifying a set of nodes with conflicting or mutually exclusive influences in multiple paths as a conflict-sensitive node group; performing conflict intensity analysis on multiple active nodes in the same target area that have opposite or excessively superimposed adjustment behaviors within the same time window based on the conflict-sensitive node group; when the interference intensity between node pairs exceeds a preset intensity threshold and the conflict entropy is greater than a set upper limit, it is marked as having an explicit behavioral conflict, and the associated node pair is added to the conflict relationship set, while limiting the behavioral amplitude or delayed response mode of low-priority nodes in the overlapping adjustment area.

[0016] In a preferred embodiment, the step of constructing a strategy optimization model based on regulation roles and conflict relationships to solve for the optimal combination of regulation strategies for each node specifically involves: constructing a heterogeneous graph of regulation roles and embedding the heterogeneous graph structure into a graph differentiable optimization framework; introducing node behavior role embedding vectors, local conflict intensity, and a global regulation objective function to construct a cross-role graph structure strategy combination space; based on the combination space, using a multi-objective differentiable combinatorial optimization method, jointly solving for the Pareto optimal strategy set for the system voltage stability objective, the regulation power loss minimization objective, and the regulation behavior consistency objective; based on the optimal strategy set, using a graph attention mechanism to compress and filter regulation influence paths to generate a low-dimensional strategy candidate set; selecting personalized strategies based on the node's historical response inertia and neighborhood cooperative matching degree; encoding the personalized strategies into parameterized regulation trajectories, deploying them end-to-end in each node using a graph neural network inference model, and performing structural updates and response retraining during operation.

[0017] In a preferred embodiment, each optical storage node performs differentiated voltage control actions according to its assigned regulation role and specific strategy. Specifically, each optical storage node estimates local voltage changes in advance within the current time window based on its corresponding role label and strategy candidate set, combined with a local voltage trend prediction model. Based on the voltage trend prediction results, it selects the regulation trajectory with the best forward performance from the strategy candidate set. Each node dynamically constructs a local collaborative control domain based on the communication connection status and voltage influence path in its neighborhood, and shares the regulation state with key nodes in the control domain to achieve edge self-organized collaborative regulation and avoid centralized control bottlenecks. The regulation trajectory is incrementally corrected by continuously monitoring the voltage response residual during the node's execution process.

[0018] The technical effects and advantages of the voltage coordination control method and system based on distributed photovoltaic energy storage in this invention are as follows:

[0019] 1. This invention introduces a trend vector and a behavioral role classification mechanism, utilizing graph neural networks and game theory modeling to accurately identify dominant, responsive, and inert nodes, and establishes a hierarchical regulatory game and role matching strategy accordingly. Through differentiated strategy allocation and role-driven regulatory response, it avoids the problem of multiple nodes "ignoring the interconnected effects and simultaneously injecting reactive power" when voltage is too high, breaking the traditional "centralized decision-making" or "blind concurrent response" model, and alleviating the "regulation-counter-regulation" vicious cycle from the source.

[0020] 2. This invention constructs a voltage interaction map and a conflict path identification mechanism. This method can identify and dynamically suppress explicit behavioral conflicts and potential interference paths during the regulation process. Combined with graph differentiability optimization and personalized deployment of strategy trajectories, it achieves local coordination, self-organizing linkage, and response correction among nodes. Especially under conditions of communication delay, external disturbances, or node response deviation, the system can rely on the neighborhood behavior perception mechanism and structural self-attention adjustment mechanism to dynamically adjust the response mode, maintaining the continuity, coordination, and intelligent adaptability of voltage control. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the voltage coordination control method based on distributed optical energy storage of the present invention.

[0022] Figure 2 This is a schematic diagram of the voltage coordination control system based on distributed optical energy storage of the present invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0024] Example 1, Figure 1 The present invention provides a voltage coordination control method based on distributed optical energy storage, comprising the following steps:

[0025] S1, obtain the first data of the node, and construct the local adjustment trend vector of each node based on the time window;

[0026] The first data includes node voltage, active and reactive power output, energy storage status, and network topology information.

[0027] The process of acquiring the first data of each node and constructing a local adjustment trend vector for each node based on a time window is as follows:

[0028] The first data of the node is obtained based on the preset time window length, and the first data within the time sliding window is subjected to linear regression to extract the trend of change.

[0029] The changing trend is enhanced, and the voltage phasor difference between the node and its upstream and downstream nodes is integrated to construct a local regulation trend vector that characterizes the node's regulation sensitivity, directionality, and intervention intensity.

[0030] It should be noted that the local adjustment trend vector reflects the adjustment direction (such as voltage boosting / deboosting, reactive power injection / reactive power absorption) and adjustment intensity of the node in the current time period, and is used to characterize its voltage regulation behavior tendency.

[0031] The local adjustment trend vector is specifically:

[0032]

[0033] in, For local adjustment trend vector, Let be the voltage at node i. Let be the active power of node i. Let be the reactive power of node i. The voltage of the upstream neighboring node, This represents the voltage of the downstream neighboring node.

[0034] S2, based on the moderation trend vector, identifies the behavioral role of each node in the moderation group and performs moderation role classification;

[0035] The process of identifying the behavioral role of each node in the regulation group based on the regulation trend vector and classifying regulation roles is as follows:

[0036] Based on electrical topology and voltage phasor sensing data, a topology graph of each distributed optical-storage node is constructed. Each node is regarded as a vertex in the graph, and the edge weights between nodes are weighted by voltage coupling sensitivity to obtain a weighted graph structure.

[0037] The local adjustment trend vector of each node and the preset response inertia index (such as the frequency of historical adjustment behavior changes and the lag response time) are used as the initial node feature vector of the graph neural network.

[0038] Based on the weighted graph structure and the initial node feature vectors, a weighted graph neural network model is constructed to explore the ability of nodes to regulate regional voltage, the regulation interaction relationship and the trend aggregation direction in the topology, and to generate the behavior role embedding vector of each node.

[0039] Based on the behavioral role embedding vector, a hierarchical adjustment game model is constructed, in which the dominant node seeks the optimal coordination direction in the upper-level strategy space, and the responsive node makes secondary responses based on the dominant behavior in the lower-level strategy. The system solves the multi-stage adjustment Nash equilibrium or Stackelberg equilibrium and generates the adjustment game score of each node.

[0040] Generate role labels containing adjustment priorities based on the adjustment game scores of the nodes.

[0041] The specific formula for calculating the voltage coupling sensitivity is as follows:

[0042]

[0043] The specific formula for calculating the adjusted game score is as follows:

[0044]

[0045]

[0046] in, Let be the coupling sensitivity of reactive power regulation at node j to the voltage at node i. Let be the voltage change value at node i. The change in reactive power injected / absorbed at node j. Let i be the adjustment game score. , and These are the preset weights. To represent the degree of influence of node i on the adjustment strategies of other nodes, The degree of consistency between node behavior and game solution. For node i in a given policy Adjustment cost function under the condition, Embed a vector for the behavior role of node i. The actual adjustment strategy for node i. To obtain the optimal response, Let be the gradient of the effect of the policy of node i on other cost functions.

[0047] Furthermore, the benefits of adjusting role divisions include:

[0048] To enhance the relevance of subsequent modeling: Different role nodes have different strategic behavior boundaries and adjustment objectives. Dominant nodes can participate in upper-level coordination games to optimize and adjust trends; follower or responsive nodes mainly adjust based on dominant behavior in lower-level strategies; if role identification is not performed, the game model needs to model each node indiscriminately, which is not only computationally complex but also obscures structural patterns and affects the overall coordination efficiency.

[0049] Optimize the hierarchical focus of conflict path identification: In the voltage disturbance spectrum, the disturbance path weights of different behavioral roles should be treated differently; the disturbances of the dominant node propagate first, and its path weight is high; the disturbances of the response node may be cooperative adjustments, and the disturbance assessment should be weakened; early identification of roles helps to build a more accurate positive disturbance / reverse cancellation path identification mechanism.

[0050] Supporting the rational allocation of differentiated control strategies: A quantitative understanding of node behavior capabilities helps to: prioritize the allocation of complex adjustment tasks to dominant nodes with fast response and strong recovery; and assign edge-buffered tasks to inert response nodes; otherwise, it is impossible to achieve the "task-capability" matching of distributed collaboration, which may lead to an imbalance of "response overload" or "insufficient intervention".

[0051] S3. Based on the partitioning results, construct the voltage interaction spectrum of the local node group, identify the node group with positive interference and reverse cancellation paths, evaluate the behavioral conflict in the regulation, and obtain the conflict relationship.

[0052] The voltage interaction map of the local node group constructed based on the partitioning results is as follows:

[0053] Based on the results of the adjustment role division, the optical storage nodes that meet the preset conditions are aggregated into a local adjustment node group;

[0054] Based on each pair of nodes within the local node group, a disturbance conduction channel mapping matrix is ​​constructed according to historical voltage monitoring data and node output change data.

[0055] Based on the disturbance conduction channel mapping matrix and combined with the node behavior trend data within the time window, a voltage mutual influence spectrum is formed.

[0056] The preset conditions specifically include strong physical proximity, frequent linkage of adjustment behaviors, and a master-slave or peer relationship between role types.

[0057] The identification of node groups with both positive interference and negative cancellation paths specifically involves:

[0058] The voltage response coupling tensor between nodes is generated by calculating the product of the voltage regulation sensitivity matrix and the regulation trend vector between each node in the local node group.

[0059] Based on this tensor, a set of directed influence paths is constructed. Paths with edge weights in the same direction and positive trend correlation are identified as positive interference paths, while paths with edge weights in opposite directions and negative trend correlation are identified as negative cancellation paths.

[0060] The behavioral conflicts in the assessment and adjustment process yield conflict relationships, specifically:

[0061] A set of nodes that have conflicting, intersecting, or mutually exclusive effects in multiple paths will be identified as a conflict-sensitive node group;

[0062] Based on the conflict-sensitive node group, the conflict intensity analysis is performed on multiple active nodes that are in the same target area and have opposite or excessively superimposed adjustment behaviors within the same time window.

[0063] When the interference intensity between node pairs exceeds the preset intensity threshold and the conflict entropy is greater than the set upper limit, it is marked as having explicit behavioral conflict, and the associated node pair is added to the conflict relationship set. At the same time, by limiting and adjusting the behavior amplitude or delay response mode of low-priority nodes in the overlapping area, the potential interference path is dynamically suppressed.

[0064] The conflict entropy is specifically:

[0065]

[0066] in, For conflict entropy, The probability distribution of conflict behavior across different node groups. A categorized index for conflict behaviors.

[0067] S4. Based on the relationship between adjustment roles and conflicts, construct a strategy optimization model to solve for the optimal combination of adjustment strategies for each node;

[0068] The aforementioned strategy optimization model, based on the relationship between adjustment roles and conflicts, is used to solve for the optimal combination of adjustment strategies for each node. Specifically:

[0069] Construct a heterogeneous graph of regulation roles, in which each distributed optical storage node is a vertex of the graph, the node type label is determined by its behavioral role, and the edge weight is jointly determined by the voltage coupling sensitivity and the historical regulation interaction tensor.

[0070] By embedding heterogeneous graph structures into a graph differentiable optimization framework, node behavior role embedding vectors, local conflict intensity, and global adjustment objective function are introduced to construct a cross-role graph structure strategy combination space.

[0071] Based on the combinatorial space, a Pareto optimal policy set is jointly solved by a multi-objective differentiable combinatorial optimization method to solve for the system voltage stability objective, the minimum regulation power loss objective, and the regulation behavior consistency objective.

[0072] Based on the optimal strategy set, the influence path of regulation is compressed and filtered through graph attention mechanism to generate a low-dimensional strategy candidate set. Personalized strategies are selected based on the historical response inertia of nodes and the degree of neighborhood cooperation matching.

[0073] Personalized strategies are encoded into parameterized adjustment trajectories and deployed end-to-end at each node using a graph neural network inference model, supporting structural updates and response retraining during operation.

[0074] In this embodiment, the graph differentiable optimization framework is a graph-based differentiable optimization method. It models the optical storage nodes as vertices in a graph structure, using voltage coupling sensitivity and conflict relationships between nodes as edge weights to form a heterogeneous graph structure. Differentiable operators are used to perform gradient propagation and back-update of policy parameters within the graph structure, thereby jointly solving for the Pareto optimal policy set under the constraints of multiple objective functions (including voltage stability, minimum regulation power loss, and consistent regulation behavior). This framework supports rapid updates and response retraining based on changes in node roles or conflict relationships during operation, enabling dynamic optimization and personalized deployment of regulation strategies.

[0075] It should be noted that the regulation influence path determines which nodes have regulatory coupling relationships or behavioral conflicts, thus defining the policy combination constraints in the optimal policy set; that is, each candidate policy in the optimal policy set must conform to the cooperation or avoidance rules in the influence path (e.g., two nodes with opposite regulatory effects and overlapping paths cannot be activated at the same time); when constructing the "policy combination graph" or "regulation interaction graph", these paths will participate in policy modeling as weights or connections of the "edges" in the graph.

[0076] This embodiment proposes a multi-role regulation policy joint inference method based on graph-structured differentiable combinatorial optimization, overcoming the problems of poor coupling between training and deployment and low response flexibility in existing reinforcement learning methods. By representing the voltage regulation problem as a heterogeneous graph structure and introducing regulation role labels, conflict indices, and policy cooperation objectives, a graph-optimized solvable space is constructed. Utilizing graph attention mechanisms and Pareto front search strategies, global consistency and dynamic adaptation to local optima of the regulation scheme are achieved, significantly improving the intelligence and robustness of the system's voltage control.

[0077] S5, each optical storage node performs differentiated voltage control actions according to its assigned regulation role and specific strategy, and automatically adjusts its response mode according to changes in the behavior of neighboring nodes.

[0078] Each optical storage node performs differentiated voltage control actions according to its assigned regulation role and specific strategy, specifically as follows:

[0079] Each photovoltaic-storage node estimates local voltage changes in advance within the current time window based on its corresponding role label and strategy candidate set, combined with the local voltage trend prediction model.

[0080] The optimal forward-looking energy adjustment trajectory is selected from the strategy candidate set based on the voltage trend prediction results.

[0081] Each node dynamically constructs a local collaborative control domain based on the communication connection status and voltage influence path within its neighborhood, and shares the regulation state with key nodes in the control domain to achieve edge self-organized collaborative regulation and avoid centralized control bottlenecks.

[0082] By continuously monitoring the voltage response residuals (the deviation between actual changes and strategy predictions) during node execution, incremental corrections are made to the adjustment trajectory, improving the matching degree between the strategy trajectory and the actual operating environment, and avoiding adjustment deviations or failures due to external disturbances. The incremental correction can be implemented using recursive least squares, Kalman filtering, or a correction model based on time series residual prediction.

[0083] The adjustment trajectory is fine-tuned based on the node's role characteristics (such as dominant, responsive, or inert). Specifically:

[0084] The dominant node responds quickly according to the adjustment trajectory, implementing active voltage boosting or reduction and active / reactive power injection / absorption;

[0085] Responsive nodes perform buffered or follow-up auxiliary adjustments based on the behavior of the dominant node and the coordination rules, adjusting the output to smooth system dynamics;

[0086] Inertial response nodes perform low-frequency, small-amplitude adjustments, undertaking stability maintenance and buffering tasks.

[0087] The automatic adjustment of the response method based on changes in the behavior of neighboring nodes is specifically as follows:

[0088] Based on the voltage coupling relationship and communication topology between nodes, a behavior monitoring window for this node and neighboring nodes is established to acquire the adjustment behavior data of neighboring nodes in real time, including adjustment direction, active / reactive power output change rate and adjustment timing.

[0089] Perform trend consistency analysis on the regulatory behavior data to determine whether there is trend convergence (synergistic enhancement) or trend conflict (behavioral adversarial).

[0090] If trend convergence exists, the adjustment range should be increased or the response time should be advanced to enhance the linkage of local adjustments.

[0091] If there is a trend conflict, a buffer suppression mechanism is adopted to reduce the adjustment action rate or introduce a response delay in order to avoid voltage fluctuations from causing oscillations.

[0092] The response inertia, behavioral volatility, and regulation stability of neighboring nodes are used as auxiliary inputs to dynamically adjust the response weight parameters and regulation sensitivity of this node.

[0093] Based on the history of behavior changes, a neighborhood dynamic behavior graph is constructed, response paths are classified and categorized, and the most influential neighboring nodes are identified as the main reference objects through a structural self-attention mechanism.

[0094] After each round of adjustment response is completed, the response offset caused by changes in the behavior of neighboring nodes is evaluated, and the behavior adjustment strategy of this node is automatically updated to adapt to environmental disturbances and the evolution of neighborhood behavior.

[0095] Example 2, Figure 2 The present invention provides a voltage coordination control system based on distributed optical energy storage, comprising the following modules:

[0096] Local trend perception module: used to acquire the first data of the node and construct the local adjustment trend vector of each node based on the time window;

[0097] Behavioral role recognition module: used to identify the behavioral role of each node in the moderation group based on the moderation trend vector, and to classify the moderation roles;

[0098] Voltage Influence Map Construction Module: This module is used to construct voltage interaction maps of local node groups based on the partitioning results, identify node groups with positive interference and negative cancellation paths, evaluate behavioral conflicts during regulation, and obtain conflict relationships.

[0099] Regulation strategy generation module: used to build a strategy optimization model based on the regulation role and conflict relationship, and solve for the optimal regulation strategy combination for each node;

[0100] Differentiated execution and adjustment module: This module enables each optical storage node to execute differentiated voltage control actions based on its assigned adjustment role and specific strategy, and to automatically adjust its response mode according to changes in the behavior of neighboring nodes.

[0101] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0102] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0103] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0104] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0105] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0106] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A voltage coordination control method based on distributed optical storage, characterized in that, The method comprises the following steps: obtaining first data of nodes and constructing a local adjustment trend vector of each node based on a time window; based on the adjustment trend vector, identifying the behavior role of each node in the adjustment group, and performing adjustment role division; based on the division result, constructing a voltage mutual influence graph of the local node group, specifically: according to the adjustment role division result, aggregating the photovoltaic storage nodes meeting the preset condition into a local adjustment node group; based on each pair of nodes in the node group, constructing a disturbance conduction channel mapping matrix according to historical voltage monitoring data and node output change data, and combining node behavior trend data within the time window to form a voltage mutual influence graph; identify the node group that has a positive interference path and a reverse offset path, specifically: generate a voltage response coupling tensor between nodes by calculating the product of the voltage adjustment sensitivity matrix and the adjustment trend vector between each node in the local node group; based on the coupling tensor, construct a directed influence path set, determine that the path with the same direction weight and positive trend correlation is the positive interference path, and the path with the opposite direction weight and negative trend correlation is the reverse offset path; evaluate the behavior conflict in the adjustment to obtain a conflict relationship, specifically: identify the node set with conflict intersection or mutual influence in multiple paths as a conflict sensitive node group; based on the conflict sensitive node group, analyze the conflict intensity of multiple active nodes in the same target area and within the same time window that have opposite or excessively overlapped adjustment behaviors; when the preset determination condition is met, mark that there is an explicit behavior conflict, and add the associated node pair to the conflict relationship set, and limit the behavior amplitude or adjustment delay response mode of the low-priority node in the overlapping area; based on the adjustment role and the conflict relationship, construct a strategy optimization model to solve the optimal adjustment strategy combination of each node; each photovoltaic storage node executes differentiated voltage control actions according to the assigned adjustment role and specific strategy, and automatically adjusts the response mode according to the behavior change of the adjacent node. 2.The distributed optical storage-based voltage coordination control method according to claim 1, characterized in that, The method comprises the following steps: obtaining first data of nodes and constructing a local adjustment trend vector of each node based on a time window; based on the adjustment trend vector, identifying the behavior role of each node in the adjustment group, and performing adjustment role division; 3. The distributed optical storage-based voltage coordination control method according to claim 2, characterized in that, based on the division result, constructing a voltage mutual influence graph of the local node group, specifically: according to the adjustment role division result, aggregating the photovoltaic storage nodes meeting the preset condition into a local adjustment node group; based on each pair of nodes in the node group, constructing a disturbance conduction channel mapping matrix according to historical voltage monitoring data and node output change data, and combining node behavior trend data within the time window to form a voltage mutual influence graph; identify the node group that has a positive interference path and a reverse offset path, specifically: generate a voltage response coupling tensor between nodes by calculating the product of the voltage adjustment sensitivity matrix and the adjustment trend vector between each node in the local node group; based on the coupling tensor, construct a directed influence path set, determine that the path with the same direction weight and positive trend correlation is the positive interference path, and the path with the opposite direction weight and negative trend correlation is the reverse offset path; evaluate the behavior conflict in the adjustment to obtain a conflict relationship, specifically: identify the node set with conflict intersection or mutual influence in multiple paths as a conflict sensitive node group; based on the conflict sensitive node group, analyze the conflict intensity of multiple active nodes in the same target area and within the same time window that have opposite or excessively overlapped adjustment behaviors; when the preset determination condition is met, mark that there is an explicit behavior conflict, and add the associated node pair to the conflict relationship set, and limit the behavior amplitude or adjustment delay response mode of the low-priority node in the overlapping area; based on the adjustment role and the conflict relationship, construct a strategy optimization model to solve the optimal adjustment strategy combination of each node; each photovoltaic storage node executes differentiated voltage control actions according to the assigned adjustment role and specific strategy, and automatically adjusts the response mode according to the behavior change of the adjacent node. The method comprises the following steps: obtaining first data of nodes and constructing a local adjustment trend vector of each node based on a time window; based on the adjustment trend vector, identifying the behavior role of each node in the adjustment group, and performing adjustment role division; based on the division result, constructing a voltage mutual influence graph of the local node group, specifically: according to the adjustment role division result, aggregating the photovoltaic storage nodes meeting the preset condition into a local adjustment node group; based on each pair of nodes in the node group, constructing a disturbance conduction channel mapping matrix according to historical voltage monitoring data and node output change data, and combining node behavior trend data within the time window to form a voltage mutual influence graph; identify the node group that has a positive interference path and a reverse offset path, specifically: generate a voltage response coupling tensor between nodes by calculating the product of the voltage adjustment sensitivity matrix and the adjustment trend vector between each node in the local node group; based on the coupling tensor, construct a directed influence path set, determine that the path with the same direction weight and positive trend correlation is the positive interference path, and the path with the opposite direction weight and negative trend correlation is the reverse offset path; evaluate the behavior conflict in the adjustment to obtain a conflict relationship, specifically: identify the node set with conflict intersection or mutual influence in multiple paths as a conflict sensitive node group; based on the conflict sensitive node group, analyze the conflict intensity of multiple active nodes in the same target area and within the same time window that have opposite or excessively overlapped adjustment behaviors; when the preset determination condition is met, mark that there is an explicit behavior conflict, and add the associated node pair to the conflict relationship set, and limit the behavior amplitude or adjustment delay response mode of the low-priority node in the overlapping area; based on the adjustment role and the conflict relationship, construct a strategy optimization model to solve the optimal adjustment strategy combination of each node; each photovoltaic storage node executes differentiated voltage control actions according to the assigned adjustment role and specific strategy, and automatically adjusts the response mode according to the behavior change of the adjacent node. Generate a role label containing adjustment priority according to the adjustment game score of the node.

4. The distributed optical storage-based voltage coordination control method according to claim 3, characterized in that, The preset determination condition is satisfied, specifically: The interference intensity between the node pairs exceeds a preset intensity threshold, and the conflict entropy is greater than a set upper limit.

5. The distributed optical storage-based voltage coordination control method according to claim 4, characterized in that, The strategy optimization model is constructed based on the adjustment role and the conflict relationship, and the optimal adjustment strategy combination of each node is solved, specifically: A heterogeneous graph of adjustment roles is constructed, and the heterogeneous graph structure is embedded into a graph differentiable optimization framework, a node behavior role embedding vector, a local conflict intensity, and a global adjustment objective function are introduced, and a cross-role graph structure strategy combination space is constructed; Based on the combination space, a multi-objective differentiable combination optimization method is used to jointly solve a Pareto optimal strategy set of the system voltage stability target, the minimum adjustment power loss target, and the adjustment behavior consistency target; Based on the optimal strategy set, the adjustment influence path is compressed and screened through a graph attention mechanism to generate a low-dimensional strategy candidate set, and an individualized strategy is selected according to the historical response inertia and the neighborhood collaborative matching degree of the node; The individualized strategy is encoded into a parameterized adjustment trajectory, and the graph neural network inference model is deployed in an end-to-end manner in each node, and the structure is updated and the response is retrained during operation.

6. The distributed optical storage-based voltage coordination control method according to claim 5, characterized in that, Each optical storage node executes a differentiated voltage control action according to the assigned adjustment role and specific strategy, specifically: Each optical storage node estimates the local voltage variation in advance within the current time window based on the corresponding role label and strategy candidate set in combination with the local voltage trend prediction model; Based on the voltage trend prediction result, an optimal forward-looking performance adjustment trajectory is selected from the strategy candidate set; Each node dynamically constructs a local collaborative control domain according to the communication connection condition and the voltage influence path in the neighborhood, and shares the adjustment state with the key nodes in the control domain to realize edge self-organization collaborative adjustment and avoid centralized control bottlenecks; The adjustment trajectory is incrementally modified by continuously monitoring the voltage response residual during node execution.

7. A system using the voltage coordination control method based on distributed optical storage according to any one of claims 1 to 6, characterized in that, The following modules are included: Local trend perception module: used to obtain node first data and construct a local adjustment trend vector for each node based on a time window; Behavior role identification module: used to identify the behavior role of each node in the adjustment group based on the adjustment trend vector, and perform adjustment role division; Voltage influence graph construction module: used to construct a voltage mutual influence graph of a local node group based on the division result, identify node groups with positive interference and reverse cancellation paths, and evaluate behavior conflicts in adjustment to obtain conflict relationships; Adjustment strategy generation module: used to construct a strategy optimization model based on the adjustment role and the conflict relationship, and solve the optimal adjustment strategy combination of each node; Differential execution and adjustment module: used for each optical storage node to execute a differentiated voltage control action according to the assigned adjustment role and specific strategy, and automatically adjust the response mode according to the behavior change of the adjacent node.

Images