Intelligent decision-making method, system and equipment for operation planning of power distribution network based on thinking chain, and medium

Abstract

The invention relates to the technical field of power distribution network operation planning, and discloses a thinking chain-based intelligent decision-making method, system and equipment for power distribution network operation planning, and a medium, and the method comprises the steps: obtaining the operation perception data of a power distribution network; performing multi-dimensional analysis on the operation situation of the power distribution network based on the operation sensing data, and constructing a multi-dimensional situation model; decomposing the overall operation planning decision into sub-operation planning decisions based on the operation planning targets and the incidence relation between the operation planning targets, and generating candidate operation planning decisions based on the multi-dimensional situation model and the sub-operation planning decisions; simulating and deducing the operation state change of the power distribution network in a simulation environment to obtain a simulation deduction result; and performing strategy evaluation on the simulation deduction result of each candidate operation planning decision based on the expected operation target of the operation of the power distribution network to obtain an optimal operation planning decision. According to the invention, the dynamic adaptability and reliability of decision planning during operation of the power distribution network are improved, and stable operation of the power distribution network is guaranteed.

Description

Technical Field

[0001] This invention relates to the field of power distribution network operation planning technology, and in particular to a method, system, device and medium for intelligent decision-making in power distribution network operation planning based on thought chain. Background Technology

[0002] In the power grid sector, distribution network operation planning decisions are crucial for the safe and efficient operation of the power system. A common existing approach is based on historical data and simple rules. However, this approach reveals significant drawbacks when facing increasingly complex distribution network structures and changing operating environments, namely a lack of dynamic adaptability. Traditional methods typically rely on fixed historical data patterns and established rules to formulate distribution network operation planning decisions, making it difficult to make flexible and reasonable adjustments in real time based on the current actual operating status of the distribution network and dynamic changes in the external environment (such as the impact of sudden weather changes on distributed energy generation, sudden increases in electricity demand, etc.). For example, when extreme weather causes a significant drop in distributed power output, or when a region experiences a sudden surge in electricity load due to special events, decision-making methods based on historical data and simple rules cannot quickly and effectively replan the distribution network's operation strategy, potentially leading to reduced power supply reliability, unreasonable allocation of power resources, and other problems that seriously affect the stable operation of the distribution network. Summary of the Invention

[0003] In view of the aforementioned existing problems, the present invention provides a method, system, device and medium for intelligent decision-making of power distribution network operation planning based on thinking chain.

[0004] Therefore, the technical problem solved by this invention is to improve the dynamic adaptability and reliability of decision-making and planning in the operation of the distribution network, and to ensure the stable operation of the distribution network.

[0005] To address the aforementioned technical problems, this invention provides the following technical solution: a smart decision-making method for distribution network operation planning based on a thought chain, comprising: acquiring real-time operation perception data of the distribution network through various monitoring devices deployed in the distribution network; performing topology analysis, load characteristic analysis, and equipment status analysis on the operation scenario of the distribution network based on the operation perception data, and constructing a multi-dimensional scenario model of distribution network operation; decomposing the overall operation planning decision of the distribution network into multiple sub-operation planning decisions based on the operation planning objectives of the distribution network and the correlation between the operation planning objectives, and generating multiple candidate operation planning decisions based on the multi-dimensional scenario model and multiple sub-operation planning decisions; simulating and deducing the changes in the operation state of the distribution network after executing each candidate operation planning decision in a simulation environment of the distribution network, and obtaining the simulation deduction result of each candidate operation planning decision; and evaluating the simulation deduction result of each candidate operation planning decision based on the expected operation objectives of the distribution network operation, to obtain the optimal operation planning decision for the distribution network operation.

[0006] As a preferred embodiment of the intelligent decision-making method for power distribution network operation planning based on the thinking chain described in this invention, the step of decomposing the overall operation planning decision of the power distribution network into multiple sub-operation planning decisions includes: constructing an association matrix based on the operation planning target, traversing each matrix element in the association matrix, and constructing a target association graph based on the matrix elements that have connection relationships with the nodes. In the target association graph, starting from any target node, a strong association cluster is constructed based on traversing all connected target nodes, and a dependency matrix is ​​constructed based on the dependency relationships between each strong association cluster. The clusters with strong associations that have no predecessor dependency are arranged as the top layer and then arranged downwards to construct a cluster-level decision-making hierarchy from the top to the bottom. Based on the cluster-level decision hierarchy and the correlation between targets within each strongly correlated cluster, the overall operation planning decision of the distribution network is decomposed to obtain multiple sub-operation planning decisions.

[0007] As a preferred embodiment of the intelligent decision-making method for distribution network operation planning based on the thinking chain described in this invention, the step of decomposing the overall operation planning decision of the distribution network into multiple sub-operation planning decisions further includes determining the connection point between the first strongly correlated cluster and the second strongly correlated cluster based on the cluster-level decision hierarchy; the first strongly correlated cluster is the cluster preceding the second strongly correlated cluster. At the junction point, the output target of the first strongly correlated cluster and the input target of the second strongly correlated cluster are linked together to obtain a cross-cluster sub-decision chain; All cross-cluster sub-decision chains are arranged in the order of the cluster-level decision hierarchy to generate sub-operation planning decisions for the overall operation planning of the distribution network.

[0008] As a preferred embodiment of the intelligent decision-making method for power distribution network operation planning based on the thinking chain described in this invention, the step of generating multiple candidate operation planning decisions based on a multi-dimensional scenario model and multiple sub-operation planning decisions includes: using the analysis results of topology, load characteristics and equipment status in the multi-dimensional scenario model as scenario nodes, using sub-operation planning decisions as sub-decision nodes, constructing a directed weighted correlation graph based on the influence relationship of sub-operation planning decisions on different scenario factors, and determining the dynamic evolution rules of the scenario based on the dynamic change law of topology, load characteristics and equipment status in the multi-dimensional scenario model; Starting from the context node, traverse the directed weighted association graph to determine the effective path from the context node to the sub-decision node; The sub-decision combination patterns are integrated to generate an initial operation planning decision, and the initial operation planning decision is expanded to obtain an expanded operation planning decision; Conflict detection is performed based on the logical relationships between the sub-decisions in the extended operation planning decision during the execution process to obtain the first objective operation planning decision and the second objective operation planning decision; Conflict resolution is performed based on the relationship between the first objective operation planning decision and the multidimensional scenario model to obtain the conflict-resolved operation planning decision. The conflict-resolved operation planning decision and the second objective operation planning decision are then integrated to obtain the candidate operation planning decision.

[0009] As a preferred embodiment of the intelligent decision-making method for distribution network operation planning based on the thinking chain described in this invention, the simulation results of each candidate operation planning decision are evaluated based on the expected operation objectives of the distribution network to obtain the optimal operation planning decision for the distribution network, including: The simulation results of each candidate operation planning decision are compared with the expected operation targets of the distribution network to construct a target deviation matrix; Based on the target deviation matrix and the simulation process of each candidate operation planning decision, a target achievement path map is generated for each expected operation target. Based on the synergistic relationship between each candidate operation planning decision in the target achievement path map numerator, a decision synergy network is constructed. Based on the changing characteristics of the distribution network operation status evolution trajectory after the execution of each candidate operation planning decision, the decision stability of each candidate operation planning decision is determined; Potential operational planning decisions are obtained by screening based on the target deviation matrix, the decision stability of each candidate operational planning decision, and the number of connecting edges in the decision collaboration network. Decision priorities are determined based on the decision stability of each potential operation planning decision and the number of connecting edges in the decision collaboration network, and the potential operation planning decision corresponding to the highest decision priority is determined as the optimal operation planning decision.

[0010] As a preferred embodiment of the intelligent decision-making method for power distribution network operation planning based on the thinking chain described in this invention, the construction of a multi-dimensional scenario model for power distribution network operation includes: traversing the direct or indirect connection paths between nodes based on node connection information from any node to obtain a basic path; traversing the basic path, determining the combination of the first and last connected paths corresponding to the current node as a ring network, and decomposing the ring network based on a preset minimum cut set to obtain multiple independent paths. Based on load information, load data of various nodes in space are integrated using time intervals to obtain a spatiotemporal distribution model of load characteristics that changes over time; future load trends are predicted to obtain the predicted load results between nodes at different times. Frequency domain analysis is performed on various operating parameters to determine the equipment state fault characteristics corresponding to the frequency and amplitude change characteristics at the time of fault occurrence; based on the equipment state fault characteristics and the equipment's historical state transitions, the degradation path of the equipment state is deduced. A multi-dimensional scenario model is constructed by fusing basic paths, independent paths, predicted load results, equipment status fault characteristics, and degradation paths.

[0011] As a preferred embodiment of the intelligent decision-making method for power distribution network operation planning based on the thinking chain described in this invention, the construction of the multi-dimensional scenario model for power distribution network operation further includes constructing an initial model framework based on each node in the basic path and the edges connecting the nodes. The node attributes of each node are determined based on the predicted load amount of each node at different times in the predicted load results. Based on independent paths and the predicted load of each node, the load transmission pressure of the edges connecting each node is determined, and the load transmission pressure is determined as the edge attribute of the edges connecting each node. For devices on nodes or edges, integrate device status fault characteristics and degradation paths into the node attributes of the corresponding node or the edge attributes of the edge. The initial model framework is updated based on the node attributes of each node and the edge attributes of each edge to obtain a multidimensional context model.

[0012] This invention provides an intelligent decision-making system for power distribution network operation planning based on thought chain.

[0013] As a preferred embodiment of the intelligent decision-making system for power distribution network operation planning based on the thinking chain described in this invention, it includes a data acquisition module for acquiring real-time operational perception data of the power distribution network through various monitoring devices deployed in the power distribution network. The scenario analysis module is used to perform topology analysis, load characteristic analysis, and equipment status analysis on the operation scenario of the distribution network based on operational perception data, and to build a multi-dimensional scenario model of distribution network operation. The decision generation module is used to decompose the overall operation planning decision of the distribution network into multiple sub-operation planning decisions based on the operation planning objectives of the distribution network and the correlation between the operation planning objectives, and to generate multiple candidate operation planning decisions based on the multi-dimensional scenario model and multiple sub-operation planning decisions. The decision simulation module is used to simulate and extrapolate the changes in the operating state of the distribution network after executing each candidate operation planning decision in the simulation environment of the distribution network, and obtain the simulation results of each candidate operation planning decision; The decision evaluation module is used to evaluate the simulation results of each candidate operation planning decision based on the expected operation goals of the distribution network, and obtain the optimal operation planning decision for the distribution network.

[0014] The present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of a smart decision-making method for power distribution network operation planning based on a thought chain.

[0015] The present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of a smart decision-making method for power distribution network operation planning based on a thought chain.

[0016] The beneficial effects of this invention are as follows: By sensing the status in real time, the latest information on the operation of the distribution network can be obtained immediately. Changes in the status of internal equipment or the external environment can be captured promptly, freeing it from the constraints of fixed patterns. Decisions can be dynamically adjusted based on real-time changes, improving the dynamic adaptability of decision-making and planning. Multiple candidate operation planning decisions are generated based on the operating context of the distribution network and the correlation between operation planning objectives. Through interrelationships, collaborative optimization of the distribution network is achieved, generating more reasonable decisions. The simulation results of each candidate operation planning decision are then evaluated in a simulation environment based on the expected operating objectives to obtain the optimal operation planning decision for the distribution network. This further ensures the rationality of the optimal operation planning decision, avoiding problems such as reduced power supply reliability and uneven resource allocation caused by unreasonable decisions, thus improving the reliability of decision-making and planning and ensuring the stable operation of the distribution network. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of a smart decision-making method for power distribution network operation planning based on a thought chain, provided as an embodiment of the present invention. Detailed Implementation

[0019] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0020] Example 1, referring to Figure 1This is the first embodiment of the present invention, which provides a smart decision-making method for power distribution network operation planning based on thought chain, including: S1: Real-time acquisition of operational perception data of the distribution network through various monitoring devices deployed in the distribution network.

[0021] S2: Based on operational perception data, perform topology analysis, load characteristic analysis, and equipment status analysis on the operation scenario of the distribution network to construct a multi-dimensional scenario model of distribution network operation.

[0022] S3: Based on the operation planning objectives of the distribution network and the correlation between the operation planning objectives, the overall operation planning decision of the distribution network is decomposed into multiple sub-operation planning decisions, and multiple candidate operation planning decisions are generated based on the multi-dimensional scenario model and multiple sub-operation planning decisions.

[0023] S4: Simulate and deduce the changes in the operating state of the distribution network after executing each candidate operation planning decision in the simulation environment of the distribution network, and obtain the simulation results of each candidate operation planning decision.

[0024] S5: Based on the expected operating objectives of the distribution network, perform strategy evaluation on the simulation results of each candidate operation planning decision to obtain the optimal operation planning decision for the distribution network.

[0025] It should be noted that, in response to the problem that traditional distribution network operation planning methods rely on historical data and fixed rules and lack dynamic adaptability, this invention proposes an intelligent decision-making method based on a thought chain. By sensing the operating status of the distribution network in real time, constructing a multi-dimensional scenario model, decomposing and generating candidate decisions, and conducting simulation and strategy evaluation, this invention significantly improves the dynamic adaptability and reliability of the distribution network in the face of complex operating environments and sudden changes, thereby effectively ensuring the stable operation of the distribution network.

[0026] Example 2 is an embodiment of the present invention. Based on the above embodiment, a smart decision-making method for power distribution network operation planning based on thinking chain is provided.

[0027] Furthermore, in this embodiment of the application, step S1 involves acquiring real-time operational sensing data of the distribution network through various monitoring devices deployed in the distribution network. The specific steps include: Various monitoring devices, such as smart meters, sensors, feeder terminal units (FTUs), and distribution automation units (DTUs), are pre-deployed in the distribution network. These devices collect operational sensing data of the distribution network at a set frequency (e.g., on a second or minute basis). This operational sensing data includes node connection information, load information, and equipment parameter information. Node connection information records the connection relationships and topology of various devices (transformers, switches, lines, etc.) in the distribution network. Load information covers the size and trend of electricity load in different areas and time periods. Equipment parameter information includes the rated voltage, current, and power of the equipment, as well as real-time operating parameters such as voltage, current, temperature, and insulation status. Therefore, these monitoring devices transmit the operational sensing data of the distribution network in real time to the operation planning system via Internet of Things (IoT) communication technologies (such as 5G, fiber optics, LoRa, etc.). The operation planning system can then obtain this operational sensing data of the distribution network in real time.

[0028] Furthermore, in this embodiment of the application, step S2 performs topology analysis, load characteristic analysis, and equipment status analysis on the operating scenario of the distribution network based on operational sensing data, and constructs a multi-dimensional scenario model of the distribution network operation. Specific steps include: The operation planning system performs topology analysis, load characteristic analysis, and equipment status analysis based on operational sensing data. In topology analysis, by parsing node connection information, a distribution network topology map is constructed, clarifying the connection relationships of each device, network structure, and power flow direction. Load characteristic analysis utilizes load information and employs methods such as time series analysis and cluster analysis to uncover the variation patterns and characteristics of load at different times (e.g., peak, flat, valley) and in different regions. Equipment status analysis combines equipment parameter information and uses fault diagnosis algorithms and health assessment models to determine the current operating status (normal, warning, fault) and remaining service life of the equipment.

[0029] The operation planning system integrates the analysis results of topology, load characteristics and equipment status to construct a multi-dimensional scenario model of distribution network operation that includes multiple dimensions of topology, load characteristics and equipment status, so as to comprehensively reflect the operation status of the distribution network, as described in steps 201 to 204.

[0030] Step 201: The operation planning system uses either Depth-First Search (DFS) or Breadth-First Search (BFS) algorithms to select any node from the distribution network node connection information as a starting point. It then traverses the direct or indirect connection paths between nodes to generate a basic path set. This basic path set includes the connection relationships of all equipment nodes in the distribution network. Further, when traversing the basic paths, for a given current node, the system determines whether it forms a closed loop with its nodes connected end-to-end. If the path starts from the current node, traverses the path, and returns to the current node, then that path is identified as the ring network structure of the current node.

[0031] The operation planning system is based on the minimum cut set theory in graph theory. It calculates the minimum edge set in a ring network that can separate the network into multiple subgraphs and decomposes the ring network into multiple independent paths.

[0032] Step 202: The operation planning system divides the collected load information into fixed time intervals (e.g., 15 minutes), integrates the load data of each node in space within the same time interval, and constructs a spatiotemporal distribution model of load characteristics. The spatiotemporal distribution model of load characteristics uses time as the horizontal axis and node position as the vertical axis to intuitively display the distribution of load in different regions and at different times.

[0033] The operation planning system uses the Long Short-Term Memory (LSTM) network algorithm to train the spatiotemporal distribution model of load characteristics, learns the time series characteristics and spatial correlation characteristics of load data, predicts the load change trend between different time nodes in the future, and obtains the predicted load results. Among them, the LSTM algorithm effectively handles the long-term dependency problem in long-sequence data through the gating mechanism, thereby improving the accuracy of load prediction.

[0034] In an optional embodiment, load forecasting can also be achieved through ARIMA. Specifically, when the load change is relatively stable, a forecasting model is established by analyzing the autocorrelation and differential stationarity of historical load data, and then directly applied to the distribution network load data to predict future load trends.

[0035] In another alternative embodiment, load forecasting can also be achieved through support vector regression, specifically by processing the nonlinear relationship of load data through a kernel function, training a model using historical load data, and predicting future load values.

[0036] Step 203: The operation planning system performs a Fast Fourier Transform (FFT) on the operating parameters (such as voltage, current, temperature, etc.) in the equipment parameter information to convert the time-domain signal into a frequency-domain signal and analyze its frequency components and amplitude changes. By comparing the frequency domain characteristics under normal operation and fault conditions, the fault characteristics of the equipment are determined, that is, the unique frequency change pattern and amplitude change characteristics when a fault occurs.

[0037] The operation planning system combines historical equipment status data and uses a Markov chain model to deduce the degradation path of the equipment from normal to different states such as warning and fault, based on the current equipment status and fault characteristics. The Markov chain model is based on the state transition probability matrix to describe the transition rules of the equipment between different states.

[0038] In an optional embodiment, equipment status fault feature extraction can also be achieved through wavelet transform. Specifically, by decomposing the equipment operating parameter signal at multiple scales, the energy distribution features of different frequency bands are extracted as fault indicators. This is suitable for non-stationary signal analysis and can capture transient fault features more flexibly.

[0039] In another optional embodiment, equipment status fault feature extraction can also be achieved through time-domain statistics. Specifically, the time-domain statistical features of equipment operating parameters (such as current and voltage), such as root mean square value, peak value, waveform factor or kurtosis, are directly calculated as fault feature indicators.

[0040] Step 204: The operation planning system integrates the basic path, independent path, predicted load results, equipment status fault characteristics and degradation path to construct a multi-dimensional scenario model, as described in steps 2041 to 2045.

[0041] Step 2041: Construct an initial model framework based on each node in the basic path and the edges connecting the nodes. The runtime planning system, based on graph theory, abstracts each node in the basic path as a vertex in a graph structure, and the connections (edges) between nodes are directly mapped to the edges of the graph, thus constructing the initial model framework of the distribution network. In this framework, nodes represent various devices in the distribution network (such as transformers, switches, substations, etc.), and edges represent the electrical connections between devices, forming a graph model that can intuitively display the topology of the distribution network.

[0042] Step 2042: Determine the node attributes for each node based on the predicted load amount of each node at different times in the predicted load results. The operation planning system assigns corresponding attributes to each node based on the predicted load amount at different times according to the predicted load results. Node attributes not only include the predicted load value, but may also include information such as load change trends (e.g., rising, falling, stable), and the time of load peak occurrence. Therefore, by binding these attributes to the node, the node can bear dynamic information related to load characteristics.

[0043] Step 2043: The operation planning system determines the load transmission pressure of an edge by calculating the load transmission capacity of each edge based on the predicted load of independent paths and nodes. In this embodiment of the invention, the power flow calculation method is used in conjunction with the electrical parameters of the distribution network (such as line resistance, reactance, etc.) to calculate the load transmission pressure under different predicted load conditions by the current magnitude of each edge. The load transmission pressure is quantified by the current magnitude or by power loss and other indicators obtained by converting the current, and the pressure value is used as the attribute of the edge. Therefore, by determining the attribute of the edge, the load carrying capacity of each connected line in the distribution network can be intuitively reflected.

[0044] Step 2044: The planning system traverses the nodes and edges in the initial model framework. For devices deployed on nodes or edges, their device state fault characteristics (such as fault frequency components, amplitude variation characteristics, etc.) and degradation paths (state change trends derived from Markov chains) are integrated into the attributes of the corresponding nodes or edges. If the device is installed on a node, the device state information is added to the node attributes; if the device is located on a connecting edge (such as a switch on a line), the information is integrated into the edge attributes.

[0045] Step 2045: The planning system fully updates the initial model framework based on the determined node attributes of each node and edge attributes of each edge. Specifically, the information in the node attributes and edge attributes is integrated into the corresponding elements of the model. Through visualization technology (such as 3D graphics display, dynamic data annotation, etc.), the model can intuitively and three-dimensionally present multi-dimensional information such as the distribution network topology, load characteristics and equipment status, and finally construct a complete multi-dimensional scenario model.

[0046] Based on the analysis results of topology, load characteristics and equipment status, a multi-dimensional scenario model is constructed. This allows for the comprehensive integration of distribution network operation information, enabling the formulation of reasonable operation planning decisions based on the analysis results of the operation scenarios. This avoids problems such as reduced power supply reliability and uneven resource allocation caused by unreasonable decisions, thereby improving the reliability of decision planning and ensuring the stable operation of the distribution network.

[0047] Furthermore, in this embodiment, step S3, based on the operation planning objectives of the distribution network and the correlation between these objectives, decomposes the overall operation planning decision of the distribution network into multiple sub-operation planning decisions, and generates multiple candidate operation planning decisions based on a multi-dimensional scenario model and the multiple sub-operation planning decisions. The specific steps include: The operation planning system clearly defines the operation planning objectives of the distribution network, such as improving power supply reliability, reducing line losses, optimizing energy distribution, and improving voltage quality, and analyzes the correlation between these objectives (e.g., improving power supply reliability may increase investment costs, and reducing line losses and optimizing energy distribution have synergistic effects). Furthermore, based on the operation planning objectives and the correlations between them, the operation planning system decomposes the overall operation planning decision of the distribution network into multiple sub-operation planning decisions, such as equipment upgrade and renovation decisions, load transfer decisions, and distributed power source access decisions, as described in steps 301 to 304.

[0048] Step 301: The operation planning system assigns numbers to the operation planning objectives of the distribution network and constructs a... The correlation matrix ( (Target quantity). Elements in the matrix. Indicates operational planning objectives With operational planning objectives The relationship between them, if the operational planning objectives With operational planning objectives If there is a relationship, then =1 and =1; if no association exists, then =0 and =0. The runtime planning system traverses each element of the association matrix, using the runtime planning target as the node and the association relationship as the edge, and constructs a target association graph by connecting the nodes. In the target association graph, the node represents the runtime planning target, and the existence of the edge indicates that there is an association between the targets.

[0049] Step 302: In the target association graph, the running planning system selects any target node as the starting point and uses the depth-first search (DFS) or breadth-first search (BFS) algorithm to traverse all target nodes that are directly or indirectly connected to the node. These closely connected target nodes are then used to construct a strong association cluster. Targets within a strong association cluster have a high degree of interrelationship. Therefore, the running planning system repeats this process to find all strong association clusters in the target association graph.

[0050] The runtime planning system analyzes the dependencies between strongly correlated clusters. If the implementation of strongly correlated cluster A depends on the completion of strongly correlated cluster B, then a dependency relationship is defined between strongly correlated cluster A and strongly correlated cluster B. Based on these dependencies, the runtime planning system constructs a... Dependency matrix ( (Number of strongly correlated clusters), matrix elements Indicates a strongly correlated cluster For strongly correlated clusters Dependencies, if strongly related clusters Depends on strongly associated clusters ,but =1, otherwise =0.

[0051] In an optional embodiment, the strong association cluster can also be constructed by manually dividing the node degree. Specifically, the degree (i.e. the number of connecting edges) of each node in the target association graph is directly calculated, and the core target nodes with the highest degree and that are connected to each other are manually divided into the same strong association cluster. Then, their adjacent nodes are included one by one to form a cluster.

[0052] In another alternative embodiment, the construction of a strongly associated cluster can also be achieved through connectivity component identification. Specifically, all mutually reachable (path-connected) nodes in the target association graph are automatically identified as a strongly associated cluster.

[0053] Step 303: The planning system, based on the dependency matrix, identifies strongly related clusters without predecessor dependencies (i.e., all elements in the corresponding column of the matrix are 0), and determines them as the top-level decision level. It then sequentially searches for the next level of strongly related clusters that depend on the top-level strongly related clusters, arranging them downwards according to the dependencies, thus constructing a cluster-level decision hierarchy structure from high to low levels. Therefore, in the hierarchy structure, the execution of high-level decisions is the foundation for the implementation of low-level decisions, clearly demonstrating the execution order and dependency logic between different strongly related clusters.

[0054] Step 304: Based on the cluster-level decision hierarchy and the correlation between targets within each strongly correlated cluster, the overall operation planning decision of the distribution network is decomposed to obtain multiple sub-operation planning decisions.

[0055] The operation planning system decomposes the overall operation planning decision of the distribution network according to the cluster-level decision hierarchy and the correlation between the targets within each strongly correlated cluster, resulting in multiple sub-operation planning decisions, as described in steps 3041 to 3044.

[0056] Step 3041: For each strongly correlated cluster, the system selects a key objective from the cluster's objectives that has a high degree of influence on other objectives and plays a core driving role as the root node. Then, based on the relationships between objectives in the objective association graph, objectives directly related to the root node are added as child nodes under the root node. If a child node is also related to other objectives, the branches are extended to construct a complete sub-decision tree. In the sub-decision tree, each branch path from the root node to a leaf node represents a sub-decision sequence, which reflects the order and logical relationship of achieving objectives within the strongly correlated cluster.

[0057] Step 3042: The planning system, based on the cluster-level decision hierarchy, sequentially determines the connection points between two adjacent strongly correlated clusters (the first strongly correlated cluster and the second strongly correlated cluster) in order from the highest to the lowest level. The connection points refer to the targets in the first strongly correlated cluster that have a direct impact on or serve as the basis for the achievement of the goal of the second strongly correlated cluster, as well as the targets in the second strongly correlated cluster that depend on the output results of the first strongly correlated cluster. Therefore, by analyzing the correlation and dependency logic of the targets within the two strongly correlated clusters, these key targets that play a connecting role are identified as connection points.

[0058] Step 3043: At the determined connection point, the planning system connects the sub-decision trees of the two strongly associated clusters according to the correlation between the output target of the first strongly associated cluster and the input target of the second strongly associated cluster. Specifically, the planning system connects the branch paths related to the connection point in the sub-decision tree of the first strongly associated cluster with the branch paths starting from the connection point in the sub-decision tree of the second strongly associated cluster, forming a sub-decision chain spanning the two strongly associated clusters. During the connection process, the logical coherence of the decision sequence and the order of target achievement are ensured.

[0059] Step 3044: The operation planning system arranges all cross-cluster sub-decision chains according to the order of the cluster-level decision hierarchy, starting from the top-level strongly correlated cluster and sequentially connecting to the cross-cluster sub-decision chains of the next strongly correlated cluster, forming a complete set of decision sequences, namely the sub-operation planning decisions of the overall operation planning decision of the distribution network. The sub-operation planning decisions cover all aspects of the distribution network operation planning and have a clear execution order and logical relationship, providing detailed guidance for the implementation of the distribution network operation planning.

[0060] Furthermore, the operation planning system combines a multi-dimensional scenario model to generate multiple candidate operation planning decisions for each sub-operation planning decision, taking into account different decision schemes and parameter settings, as described in steps 305 to 309.

[0061] Step 305: The operation planning system abstracts the analysis results of topology, load characteristics, and equipment status in the multi-dimensional scenario model into scenario nodes. For example, the topological connection relationship of a certain section of the distribution network, the load curve characteristics of a certain area, and the fault characteristics of a certain transformer are taken as independent scenario nodes. At the same time, sub-operation planning decisions are taken as sub-decision nodes. Based on the degree of influence of sub-operation planning decisions on different scenario factors, the weight of the correlation is determined by calculating the influence factors, and a directed weighted correlation graph is constructed. The influence factors can be obtained based on historical data and the implementation effect of decisions, using methods such as regression analysis.

[0062] The operation planning system analyzes the patterns of topology, load characteristics, and equipment status changes over time or under other conditions in a multi-dimensional scenario model, and formulates corresponding dynamic evolution rules for the scenarios, such as load variation patterns in different seasons and the patterns of status changes caused by equipment aging.

[0063] Step 306: The planning system traverses the directed weighted association graph, starting from the context node. This embodiment employs a graph traversal algorithm (such as Depth-First Search (DFS) or Breadth-First Search (BFS)) to find paths leading to sub-decision nodes along directed edges, starting from the initial context node. During the traversal, valid paths are selected based on the weights of the association graph and preset rules. Valid paths must meet certain conditions, such as the total weight of the path reaching a certain threshold, or the sub-decision combinations involved in the path being logically reasonable and practically feasible. Valid paths represent sub-decision combination patterns driven by specific contextual factors, clearly reflecting the relationship between contextual factors and sub-decisions.

[0064] In an optional embodiment, the selection of effective paths can also be achieved through finite path search. Specifically, starting from the initial scenario node, at each step, only the edge with the highest weight is selected for extension until the sub-decision node is reached, thus forming a locally optimal effective path. Multiple such paths can be generated by changing different starting nodes or randomly selecting some nodes as starting points.

[0065] In another alternative embodiment, the selection of valid paths can also be achieved through random walks. Specifically, starting from the initial scenario node, a random walk is performed in the association graph (i.e., a random outgoing edge is selected at each step), and all paths that eventually reach the sub-decision node are recorded. Then, from these randomly generated paths, paths whose total weight exceeds a preset threshold are selected as valid paths.

[0066] Step 307: The operation planning system integrates the sub-decision combination patterns represented by the effective path based on preset decision logic (such as goal-oriented logic, priority logic, etc.), arranges the sub-decision in a certain order, and generates the initial operation planning decision.

[0067] The operation planning system expands the initial operation planning decision based on the dynamic evolution rules of the scenario and considering possible changes in scenario factors in the future. During the expansion process, sub-decision to cope with scenario changes is added or the execution order and parameters of existing sub-decision are adjusted to obtain the expanded operation planning decision, which can better adapt to the dynamic changes in the distribution network operation scenario.

[0068] Step 308: The operation planning system analyzes the logical relationships between the sub-decision in the extended operation planning decision during execution to determine if conflicts exist. By establishing a logical conflict detection model, it checks for time conflicts (e.g., two sub-decision cannot be executed simultaneously), resource conflicts (e.g., the resources required to execute sub-decisions contradict each other), and goal conflicts (e.g., the effects of sub-decision implementation cancel each other out). Based on the detection results, operation planning decisions with logical conflicts are marked as first-goal operation planning decisions, and those without logical conflicts are marked as second-goal operation planning decisions.

[0069] Step 309: For the first objective's operational planning decision, the operational planning system analyzes its relationship with the multidimensional scenario model to identify the scenario factors and decision factors leading to conflict. Conflict is resolved by adjusting the parameters of sub-decision, changing the execution order, or introducing new sub-decision, resulting in a conflict-resolved operational planning decision. The operational planning system integrates the conflict-resolved operational planning decision with the second objective's operational planning decision, removing duplicates and merging similar decisions to finally obtain candidate operational planning decisions.

[0070] Furthermore, in this embodiment of the application, step S4 simulates and extrapolates the changes in the operating state of the distribution network after executing each candidate operation planning decision in a distribution network simulation environment, and obtains the simulation and extrapolation results of each candidate operation planning decision. The specific steps include: In this embodiment, a simulation environment for the distribution network is pre-built in the operation planning system. The simulation environment is based on the actual parameters and model of the distribution network and can realistically simulate the operation of the distribution network. Therefore, the operation planning system inputs each candidate operation planning decision into the simulation environment and adjusts the operation parameters of the distribution network according to each candidate operation planning decision (such as changes in equipment parameters after equipment replacement, changes in load in each area after load transfer, changes in power injection after distributed power source access, etc.).

[0071] The operation planning system simulates the changes in the operating status of the distribution network after adjustment through simulation algorithms, calculates the changes of a series of operating indicators (such as node voltage, line current, power supply reliability indicators, line loss rate, etc.) over time, and obtains the simulation results of each candidate operation planning decision.

[0072] Furthermore, in this embodiment, step S5 evaluates the simulation results of each candidate operation planning decision based on the expected operation objectives of the distribution network, and obtains the optimal operation planning decision for the distribution network. The specific steps include: The expected operating targets for the distribution network are set in the operation planning system. Therefore, the operation planning system evaluates the simulation results of each candidate operation planning decision based on the expected operating targets to obtain the optimal operation planning decision for the distribution network, as shown in steps 501 to 505.

[0073] Step 501: The operation planning system compares the simulation results of each candidate operation planning decision with the expected operation goals of the distribution network (such as improving power supply reliability, reducing line losses, and optimizing energy allocation). For each expected operation goal, the deviation between the simulation results of the candidate operation planning decision and the target value is calculated. The deviation can be calculated using methods such as absolute error and relative error. For example, there are... Candidate operational planning decisions and If there is an expected operating goal, then build a Target deviation matrix , where matrix elements Indicates the first The candidate operation planning decision is made in the first The deviation values ​​under the expected operational objectives are presented in a matrix. This matrix visually displays the degree of deviation of each candidate operational plan decision from different objective dimensions.

[0074] Step 502: The operation planning system generates a target achievement path diagram for each expected operational target based on the target deviation matrix and the simulation process of candidate operation planning decisions. In the target achievement path diagram, candidate operation planning decisions are used as nodes to display the decision implementation path from the current state to achieving the expected target. By analyzing the correlation and collaboration relationships of different candidate operation planning decisions in each target achievement path, the synergistic relationship between candidate operation planning decisions is determined. If two candidate operation planning decisions work together and mutually promote each other in multiple target achievement paths, they are considered to have a synergistic relationship. Further, the operation planning system constructs a decision collaboration network based on the synergistic relationship, with nodes representing candidate operation planning decisions and edges representing synergistic relationships. The weight of the edges can be quantified according to the degree of synergy (e.g., calculated based on the number of jointly acting targets or the intensity of the synergistic effect).

[0075] Step 503: The operation planning system analyzes the characteristics of the distribution network operation status evolution trajectory after the execution of each candidate operation planning decision, including the fluctuation amplitude, frequency of change, and presence of abrupt changes in operating indicators (such as voltage, current, and power). The stability of the operation status is quantified by calculating statistical quantities such as variance and coefficient of variation of the evolution trajectory. A smaller variance or coefficient of variation indicates a more stable distribution network operation status after decision execution, meaning a higher decision stability for the candidate operation planning decision. For example, candidate operation planning decisions... The time series data of a key operating indicator of the distribution network after execution are as follows: Then the formula for calculating its variance is: in, The variance is the mean of the time series. Therefore, in this embodiment of the invention, the decision stability ranking of each candidate operation planning decision is determined based on the variance.

[0076] Step 504: The operation planning system comprehensively considers the target deviation matrix, decision stability, and the number of connecting edges in the decision collaboration network to screen candidate operation planning decisions. In this embodiment of the invention, candidate operation planning decisions with smaller overall deviation in the target deviation matrix, higher decision stability, and a larger number of connecting edges in the decision collaboration network (indicating good collaboration) are selected as potential operation planning decisions. For example, one screening rule for potential operation planning decisions may be: the sum of all elements in the target deviation matrix is ​​less than a certain threshold; the decision stability ranks among the top several proportions; and the number of connecting edges in the decision collaboration network is greater than a certain number.

[0077] Step 505: The operation planning system determines the decision priority based on the decision stability of each potential operation planning decision and the number of connecting edges in the decision collaboration network. A comprehensive scoring method can be used, assigning different weights to decision stability and the number of connecting edges (e.g., decision stability weight 0.6, connecting edge weight 0.4), and calculating the comprehensive score for each potential decision. The stability score is obtained by normalizing the variance ranking; the edge count score is obtained by normalizing the actual number of edges.

[0078] The operation planning system identifies the operation planning decision with the highest overall score as the optimal operation planning decision.

[0079] By evaluating the simulation results of each candidate operation planning decision in the simulation environment based on the expected operation objectives, the optimal operation planning decision for the distribution network is obtained. This further ensures the rationality of the optimal operation planning decision, avoids problems such as reduced power supply reliability and uneven resource allocation caused by unreasonable decisions, improves the reliability of decision planning, and ensures the stable operation of the distribution network.

[0080] Example 3 is the third embodiment of the present invention, which differs from the previous two embodiments in that: This embodiment also provides a distribution network operation planning intelligent decision-making system based on thought chain, including: The data acquisition module is used to acquire real-time operational sensing data of the power distribution network through various monitoring devices deployed in the power distribution network; The scenario analysis module is used to perform topology analysis, load characteristic analysis, and equipment status analysis on the operation scenario of the distribution network based on operational perception data, and to build a multi-dimensional scenario model of distribution network operation. The decision generation module is used to decompose the overall operation planning decision of the distribution network into multiple sub-operation planning decisions based on the operation planning objectives of the distribution network and the correlation between the operation planning objectives, and to generate multiple candidate operation planning decisions based on the multi-dimensional scenario model and multiple sub-operation planning decisions. The decision simulation module is used to simulate and extrapolate the changes in the operating state of the distribution network after executing each candidate operation planning decision in the simulation environment of the distribution network, and obtain the simulation results of each candidate operation planning decision; The decision evaluation module is used to evaluate the simulation results of each candidate operation planning decision based on the expected operation goals of the distribution network, and obtain the optimal operation planning decision for the distribution network.

[0081] This embodiment also provides an electronic device, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the intelligent decision-making method for power distribution network operation planning based on the thought chain proposed in the above embodiment.

[0082] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements a smart decision-making method for power distribution network operation planning based on the thought chain proposed in the above embodiment.

[0083] The storage medium proposed in this embodiment belongs to the same inventive concept as the intelligent decision-making method for power distribution network operation planning based on the above embodiment. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0084] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0085] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A smart decision-making method for distribution network operation planning based on thought chain, characterized in that: include, Real-time operational data of the power distribution network is obtained through various monitoring devices deployed in the power distribution network. Based on operational perception data, topology analysis, load characteristic analysis, and equipment status analysis are performed on the operation scenarios of the distribution network to construct a multi-dimensional scenario model of distribution network operation; Based on the operation planning objectives of the distribution network and the correlation between them, the overall operation planning decision of the distribution network is decomposed into multiple sub-operation planning decisions, and multiple candidate operation planning decisions are generated based on the multi-dimensional scenario model and multiple sub-operation planning decisions. The changes in the operating state of the distribution network after each candidate operation planning decision are simulated and deduced in the simulation environment of the distribution network, and the simulation results of each candidate operation planning decision are obtained. Based on the expected operational goals of the distribution network, the simulation results of each candidate operation planning decision are evaluated to obtain the optimal operation planning decision for the distribution network.

2. The intelligent decision-making method for distribution network operation planning based on thinking chain as described in claim 1, characterized in that: The step of decomposing the overall operation planning decision of the distribution network into multiple sub-operation planning decisions includes: constructing an association matrix based on the operation planning objectives, traversing each matrix element in the association matrix, and constructing a target association graph based on the matrix elements that have connection relationships with the nodes. In the target association graph, starting from any target node, a strong association cluster is constructed based on traversing all connected target nodes, and a dependency matrix is ​​constructed based on the dependency relationships between each strong association cluster. The clusters with strong associations that have no predecessor dependency are arranged as the top layer and then arranged downwards to construct a cluster-level decision-making hierarchy from the top to the bottom. Based on the cluster-level decision hierarchy and the correlation between targets within each strongly correlated cluster, the overall operation planning decision of the distribution network is decomposed to obtain multiple sub-operation planning decisions.

3. The intelligent decision-making method for distribution network operation planning based on thinking chain as described in claim 2, characterized in that: The process of decomposing the overall operation planning decision of the distribution network into multiple sub-operation planning decisions also includes, The connection point between the first and second strongly correlated clusters is determined based on the cluster-level decision hierarchy. The first strongly correlated cluster is the cluster preceding the second strongly correlated cluster; At the junction point, the output target of the first strongly correlated cluster and the input target of the second strongly correlated cluster are linked together to obtain a cross-cluster sub-decision chain; All cross-cluster sub-decision chains are arranged in the order of the cluster-level decision hierarchy to generate sub-operation planning decisions for the overall operation planning of the distribution network.

4. The intelligent decision-making method for distribution network operation planning based on thinking chain as described in claim 3, characterized in that: The generation of multiple candidate operation planning decisions based on a multidimensional scenario model and multiple sub-operation planning decisions includes: using the analysis results of topology, load characteristics and equipment status in the multidimensional scenario model as scenario nodes, using sub-operation planning decisions as sub-decision nodes, constructing a directed weighted correlation graph based on the influence relationship of sub-operation planning decisions on different scenario factors, and determining the dynamic evolution rules of the scenario based on the dynamic change law of topology, load characteristics and equipment status in the multidimensional scenario model. Starting from the context node, traverse the directed weighted association graph to determine the effective path from the context node to the sub-decision node; The sub-decision combination patterns are integrated to generate an initial operation planning decision, and the initial operation planning decision is expanded to obtain an expanded operation planning decision; Conflict detection is performed based on the logical relationships between the sub-decisions in the extended operation planning decision during the execution process to obtain the first objective operation planning decision and the second objective operation planning decision; Conflict resolution is performed based on the correlation between the first objective operation planning decision and the multidimensional scenario model to obtain the conflict-resolved operation planning decision. The conflict-resolved operation planning decision and the second objective operation planning decision are then integrated to obtain the candidate operation planning decision.

5. The intelligent decision-making method for distribution network operation planning based on thinking chain as described in claim 4, characterized in that: The simulation results of each candidate operation planning decision are used to evaluate strategies based on the expected operation objectives of the distribution network, so as to obtain the optimal operation planning decision for the distribution network, including: The simulation results of each candidate operation planning decision are compared with the expected operation targets of the distribution network to construct a target deviation matrix; Based on the target deviation matrix and the simulation process of each candidate operation planning decision, a target achievement path map is generated for each expected operation target. Based on the synergistic relationship between each candidate operation planning decision in the target achievement path map numerator, a decision synergy network is constructed. Based on the changing characteristics of the distribution network operation status evolution trajectory after the execution of each candidate operation planning decision, the decision stability of each candidate operation planning decision is determined; Potential operational planning decisions are obtained by screening based on the target deviation matrix, the decision stability of each candidate operational planning decision, and the number of connecting edges in the decision collaboration network. Decision priorities are determined based on the decision stability of each potential operation planning decision and the number of connecting edges in the decision collaboration network, and the potential operation planning decision corresponding to the highest decision priority is determined as the optimal operation planning decision.

6. The intelligent decision-making method for distribution network operation planning based on thinking chain as described in claim 5, characterized in that: The construction of the multi-dimensional scenario model for power distribution network operation includes: Starting from any node, the direct or indirect connection paths between nodes are traversed to obtain the basic path. The basic path is traversed, and the path combination of the first and last connections corresponding to the current node is determined as a ring network. The ring network is then decomposed based on the preset minimum cut set to obtain multiple independent paths. Based on load information, load data of various nodes in space are integrated using time intervals to obtain a spatiotemporal distribution model of load characteristics that changes over time; future load trends are predicted to obtain the predicted load results between nodes at different times. Frequency domain analysis is performed on various operating parameters to determine the equipment state fault characteristics corresponding to the frequency and amplitude change characteristics at the time of fault occurrence; based on the equipment state fault characteristics and the equipment's historical state transitions, the degradation path of the equipment state is deduced. A multi-dimensional scenario model is constructed by fusing basic paths, independent paths, predicted load results, equipment status fault characteristics, and degradation paths.

7. The intelligent decision-making method for distribution network operation planning based on thinking chain as described in claim 6, characterized in that: The construction of a multi-dimensional scenario model for power distribution network operation also includes building an initial model framework based on each node in the basic path and the edges connecting the nodes. The node attributes of each node are determined based on the predicted load amount of each node at different times in the predicted load results. Based on independent paths and the predicted load of each node, the load transmission pressure of the edges connecting each node is determined, and the load transmission pressure is determined as the edge attribute of the edges connecting each node. For devices on nodes or edges, integrate device status fault characteristics and degradation paths into the node attributes of the corresponding node or the edge attributes of the edge. The initial model framework is updated based on the node attributes of each node and the edge attributes of each edge to obtain a multidimensional context model.

8. A distribution network operation planning intelligent decision-making system based on thought chain, applying the distribution network operation planning intelligent decision-making method based on thought chain as described in any one of claims 1 to 7, characterized in that, include: The data acquisition module is used to acquire real-time operational sensing data of the power distribution network through various monitoring devices deployed in the power distribution network; The scenario analysis module is used to perform topology analysis, load characteristic analysis, and equipment status analysis on the operation scenario of the distribution network based on operational perception data, and to build a multi-dimensional scenario model of distribution network operation. The decision generation module is used to decompose the overall operation planning decision of the distribution network into multiple sub-operation planning decisions based on the operation planning objectives of the distribution network and the correlation between the operation planning objectives, and to generate multiple candidate operation planning decisions based on the multi-dimensional scenario model and multiple sub-operation planning decisions. The decision simulation module is used to simulate and extrapolate the changes in the operating state of the distribution network after executing each candidate operation planning decision in the simulation environment of the distribution network, and obtain the simulation results of each candidate operation planning decision; The decision evaluation module is used to evaluate the simulation results of each candidate operation planning decision based on the expected operation goals of the distribution network, and obtain the optimal operation planning decision for the distribution network.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the intelligent decision-making method for power distribution network operation planning based on any one of claims 1 to 7.

10. 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 steps of the intelligent decision-making method for power distribution network operation planning based on any one of claims 1 to 7.

Images