Industrial park power supply and distribution intelligent operation and maintenance management and control method based on industrial interconnection architecture

Abstract

The invention relates to the technical field of security defense and intelligent scheduling of power supply, and discloses an industrial park power supply and distribution intelligent operation and maintenance management and control method based on an industrial interconnection architecture, and the method comprises the steps: dynamically constructing a power supply and distribution path diagram and marking a potential disturbance propagation path through combining a physical connection relation and an operation state between nodes; and the generation of a disturbance propagation atlas and the deployment of edge nodes are realized. Power-off propagation risk grade evaluation is carried out through an edge control node based on a disturbance propagation atlas and real-time data, intelligent discrimination and strategy response of power supply and distribution paths in a region are realized, the precision and timeliness of abnormal response and the decision efficiency of power-off isolation and path switching can be effectively improved, and the reliability of power supply and distribution is improved. And the local autonomous capability and fault-tolerant robustness are enhanced, and energy consumption waste and service interruption caused by sudden power disturbance are remarkably reduced.

Description

Technical Field

[0001] The present invention relates to the field of security defense and intelligent scheduling technology for power supply, and in particular to a smart operation and maintenance control method for power supply and distribution in an industrial park based on an industrial interconnection architecture. Background Art

[0002] With the widespread deployment of industrial interconnection architectures in industrial parks, power supply and distribution systems are gradually developing towards digitalization, networking, and intelligence. Various power nodes achieve information exchange and coordinated scheduling through sensing terminals, communication networks, and control platforms. However, when sudden power outages or power disturbances occur, they can easily trigger chain reactions, causing cascading power outages in local areas or even across the entire park. Existing technologies usually rely on fixed fault-tolerant strategies or manual intervention to solve these problems, making it difficult to achieve precise isolation and adaptive switching. Furthermore, there are deficiencies in multi-source node collaboration and real-time response of edge devices, leading to problems such as delayed response, low utilization of backup paths, and reduced energy efficiency. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide an intelligent operation and maintenance control method for power supply and distribution in industrial parks based on industrial interconnection architecture, so as to solve the technical problems that the existing technology cannot effectively guarantee the power supply security caused by sudden power outages or power disturbances, as well as the defense and regulation after control of power supply and distribution in industrial parks with industrial interconnection architecture.

[0004] The present invention discloses a method for intelligent operation and maintenance of power supply and distribution in an industrial park based on an industrial interconnection architecture, the method comprising:

[0005] Based on the physical connection relationship of each power node and the power supply and distribution hierarchy information, a power supply and distribution path diagram between nodes is generated;

[0006] Based on the power supply and distribution path diagram and the historical operation data and real-time status data of each node, identifying the potential power outage propagation path of each node;

[0007] Based on the power supply and distribution path diagram, the potential power outage propagation path of each node and the corresponding propagation intensity index are marked as a disturbance propagation map;

[0008] The disturbance propagation map is deployed to each edge control node, and the edge control node dynamically evaluates the power outage propagation risk level of the power supply and distribution path in the area under its jurisdiction based on the disturbance propagation map and the real-time monitored current, voltage and node load status data in the area under its jurisdiction. The power outage propagation risk level is used to determine whether to trigger path isolation and switching operations.

[0009] Furthermore, the method further comprises:

[0010] Determining whether the power outage propagation risk level exceeds a preset threshold;

[0011] The power supply and distribution path with a positive judgment result is regarded as a high-risk transmission path, and the path isolation operation is performed through the edge control node in the jurisdiction of the high-risk transmission path;

[0012] Based on the power supply and distribution path diagram and the candidate list of backup power supply and distribution paths, the path switching operation is performed according to the preset tolerance strategy.

[0013] Furthermore, the path isolation operation includes:

[0014] A disturbance propagation directed graph is constructed based on the disturbance propagation graph, and the minimum propagation cost path algorithm is used to calculate multiple possible propagation paths from the disturbance source node to the target load area.

[0015] The propagation cost is determined based on the historical disturbance intensity, voltage fluctuation frequency and response delay coefficient between nodes;

[0016] Among all the propagation paths, the path with the lowest propagation cost is selected as the priority blocking path.

[0017] Furthermore, before the edge control node performs a power outage propagation risk level assessment on the power supply and distribution paths in the area under its jurisdiction, the method further includes:

[0018] The edge control node identifies highly sensitive nodes within its jurisdiction based on the propagation intensity index, voltage fluctuation frequency, and historical power outage response records of each node in the disturbance propagation map. These highly sensitive nodes are given higher weights in the subsequent real-time monitoring process and serve as key decision nodes in determining the power outage propagation risk level.

[0019] After determining the high-risk propagation path, when there is an abnormal disturbance signal at a highly sensitive node, the highly sensitive node is prioritized as the starting point of the high-risk propagation path to trigger the path isolation operation.

[0020] Furthermore, performing the path switching operation according to the preset tolerance strategy includes:

[0021] The pre-trained tolerance scoring model is called through the edge control node, and the node current fluctuation threshold, voltage anomaly duration, communication packet loss rate and load change rate obtained from the current monitoring are used as model inputs. The tolerance scoring result is output through the tolerance scoring model, and it is determined whether to trigger the backup power supply and distribution path switching operation based on the tolerance scoring result.

[0022] Furthermore, the candidate list of backup power supply and distribution paths includes priority information of each backup power supply and distribution path; the priority information is determined based on the current load state, average switching delay and voltage fluctuation value of the power supply and distribution path;

[0023] After the backup power supply and distribution path switching operation is triggered, the target backup power supply and distribution path is selected according to the priority.

[0024] Furthermore, when there is no valid path that meets the conditions in the candidate list of backup power supply and distribution paths, the method further includes:

[0025] The edge control node performs virtual subnet division and reconstruction based on the current power supply and distribution path topology to reconstruct the power supply and distribution area boundary.

[0026] Furthermore, the virtual subnet division and reconstruction operation includes:

[0027] Dynamically divide multiple power supply and distribution virtual subnets based on the current load level of each node, power flow stability and historical interconnection risk indicators, power supply and distribution path topology, and disturbance propagation map;

[0028] When it is detected that the risk level of a node in a certain subnet continues to increase, the boundary structure of the virtual subnet is automatically adjusted, and the power supply and distribution paths of the boundary nodes are reconfigured.

[0029] Furthermore, the dynamic partitioning of the virtual subnet is performed based on a partition scoring function; the scoring function calculates the association weights between nodes based on the average load value of each node, the power supply and distribution coupling strength, the historical interconnection fluctuation rate, and the consistency of power flow direction;

[0030] The edge control node clusters the node set in the power supply and distribution path diagram using a clustering algorithm based on the scoring function result, and adjusts the power supply and distribution control affiliation and the switching boundary priority strategy of each node according to the clustering result.

[0031] Furthermore, the method further comprises:

[0032] After completing the path isolation and backup power supply and distribution path switching operations, record the execution status information of this round of operations and upload the execution status information to the dispatch center;

[0033] The dispatch center dynamically modifies the indicator data in the disturbance propagation map based on the accumulated feedback data from multiple rounds of execution.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] This invention dynamically constructs a power supply and distribution path map and labels potential disturbance propagation paths by combining the physical connection relationships and operating status of each node, enabling the generation of a disturbance propagation map and the deployment of edge nodes. Edge control nodes assess the risk level of power outage propagation based on the disturbance propagation map and real-time data, enabling intelligent identification and strategic response of power supply and distribution paths within the region. This not only effectively improves the accuracy and timeliness of abnormal responses, as well as the efficiency of decision-making for power outage isolation and path switching, but also enhances local autonomy and fault tolerance, significantly reducing energy waste and service interruptions caused by sudden power disturbances. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of the application, and do not constitute a limitation of the embodiments of the present invention. In the drawings:

[0037] Figure 1 This is a flow chart of a method for intelligent operation, maintenance and control of power supply and distribution in an industrial park based on an industrial interconnection architecture disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0038] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0039] Example 1

[0040] The first aspect of the present invention discloses a method for intelligent operation and maintenance of power supply and distribution in industrial parks based on industrial interconnection architecture, which relates to the fields of security defense and intelligent dispatching technology of power supply; please refer to Figure 1 , Figure 1 This is a flow chart of a method for intelligent operation and maintenance of power supply and distribution in an industrial park based on an industrial interconnection architecture disclosed in an embodiment of the present invention. The method includes:

[0041] Based on the physical connection relationship of each power node and the power supply and distribution hierarchy information, a power supply and distribution path diagram between nodes is generated;

[0042] Based on the power supply and distribution path diagram and the historical operation data and real-time status data of each node, identifying the potential power outage propagation path of each node;

[0043] Based on the power supply and distribution path diagram, the potential power outage propagation path of each node and the corresponding propagation intensity index are marked as a disturbance propagation map;

[0044] The disturbance propagation map is deployed to each edge control node, and the edge control node dynamically evaluates the power outage propagation risk level of the power supply and distribution path in the area under its jurisdiction based on the disturbance propagation map and the real-time monitored current, voltage and node load status data in the area under its jurisdiction. The power outage propagation risk level is used to determine whether to trigger path isolation and switching operations.

[0045] Furthermore, the method further comprises:

[0046] Determining whether the power outage propagation risk level exceeds a preset threshold;

[0047] The power supply and distribution path with a positive judgment result is regarded as a high-risk transmission path, and the path isolation operation is performed through the edge control node in the jurisdiction of the high-risk transmission path;

[0048] Based on the power supply and distribution path diagram and the candidate list of backup power supply and distribution paths, the path switching operation is performed according to the preset tolerance strategy.

[0049] Specifically, in an embodiment of the present invention, the construction operation of the power supply and distribution path diagram is first based on the physical connection relationship and voltage level and other power supply and distribution hierarchical information between various power nodes in the industrial park (including substations, distribution cabinets, switch stations, key load nodes, etc.), and performs a topology construction operation to generate a complete power supply and distribution network map. The power supply and distribution path diagram indicates the set of available power supply and distribution paths from the upper-level power supply and distribution node to the lower-level load node, and includes the connection edges between nodes, the power supply and distribution direction, and the electrical dependency between nodes. Each path in the path diagram not only represents the logical path for power transmission, but also records the historical operating stability, node connection strength, path depth and other structural attributes of the path.

[0050] Based on the power supply and distribution path diagram, the present invention further provides a potential outage propagation path identification operation. A disturbance impact model is constructed based on the node topology and historical node operation data in the power supply and distribution path diagram. Specifically, based on historical outage event records and node operation log data, statistical features such as response delay, impact relationship, and coordinated fluctuation frequency between nodes are extracted to construct an initial propagation probability matrix. This matrix reflects the probability and direction of the disturbance propagation from the source node to downstream nodes.

[0051] On this basis, a graph neural network structure is introduced to iteratively train the above probability matrix. Using the physical topology of the power supply and distribution network as its foundational graph, the graph neural network integrates multiple sources of information, including power flow directionality, electrical coupling strength, operating status, and historical co-fluctuation patterns at each node. It gradually optimizes the disturbance propagation relationships between nodes and ultimately outputs a propagation intensity index matrix. This index reflects the risk level of disturbance propagation along each path, and its value can be expressed as a propagation probability, propagation delay value, or a discrete propagation risk level.

[0052] Based on this model, the propagation strength index of the disturbance path is dynamically modified by combining the voltage fluctuation value, load change rate, and communication stability parameters currently monitored by the node in real time. This allows the identification of potential power outage propagation paths for each node. A potential power outage propagation path represents the set of paths that could potentially cause a cascading power outage under certain conditions, starting from the source node and following the directed edges in the power supply and distribution path graph.

[0053] After identifying potential outage propagation paths, the results are mapped back to the original power supply and distribution path diagram. Based on the output of the disturbance impact model, each propagation path is assigned a propagation strength index, ultimately constructing a disturbance propagation map. This disturbance propagation map superimposes a graph structure representing dynamic propagation characteristics on top of the static power supply and distribution path diagram. Its core function is to support subsequent risk identification, path isolation control, and backup path scheduling.

[0054] As a further preferred embodiment, the disturbance propagation map may include but is not limited to annotation information such as node sensitivity level, historical propagation event path records, and propagation strength index. Among them, the node sensitivity level is used to measure the susceptibility of each node in the disturbance propagation process, and is comprehensively evaluated and assigned based on factors such as historical power outage first node statistics, node voltage fluctuation rate, load instability, and propagation trigger frequency. The historical propagation event path records are extracted from the preset fault event log library, recording the actual propagation links and power outage diffusion process, which are used to support the historical verification, reasoning correction and path confidence enhancement of the disturbance propagation map. The recorded information mainly comes from the abnormal event reporting data of the edge control node and the propagation chain archiving results of the central control platform.

[0055] By integrating the above three types of annotations, the disturbance propagation map not only presents static network structure information, but also reflects the dynamic response relationship under the impact of multi-source disturbances. It can serve as a key input basis for subsequent path evaluation, virtual subnet division and scheduling strategy determination.

[0056] Furthermore, after generating a disturbance propagation map, it is deployed in a structured manner to each edge control node. Each edge control node is responsible for monitoring the power supply and distribution status and assessing risks within its jurisdiction. The node sensitivity level, propagation intensity index, and historical propagation records in the disturbance propagation map serve as important a priori basis for edge control nodes to determine the propagation risk level.

[0057] During implementation, edge control nodes continuously receive real-time status data from their power nodes, including key indicators such as current amplitude, voltage fluctuation, duration of voltage anomalies, and instantaneous load change rate. They normalize this real-time data and compare it with historical operating thresholds to determine whether the current state has deviated from the normal operating range. If a key status indicator reaches the warning threshold, the node will be marked as a candidate disturbance source.

[0058] The edge control node then combines the propagation path structure modeled in the disturbance propagation map with the propagation strength indicator matrix, starting from the currently detected candidate disturbance source node, to deduce possible propagation paths. During path deduction, a comprehensive consideration of the propagation strength value, path length, the current load level of the nodes along the path, and power fluctuation trends is used to calculate the propagation risk score for each propagation path. This calculation is performed using a weighted function.

[0059] After the assessment is complete, the propagation risk score is compared with pre-set multi-level risk thresholds to output the power outage propagation risk level for each potential propagation path in the current area, optionally categorized as high risk, medium risk, and low risk. For propagation paths assessed as high risk, backup path prediction or path isolation strategies are further triggered, providing a decision-making basis for improving regional power supply resilience and reducing the spread of disturbances.

[0060] Through the above operations, the edge control node can realize distributed real-time power outage risk perception based on the disturbance propagation map, significantly improving the dynamic risk response capability and local autonomous control level of the campus power supply and distribution system.

[0061] Furthermore, the path isolation operation includes:

[0062] A disturbance propagation directed graph is constructed based on the disturbance propagation graph, and the minimum propagation cost path algorithm is used to calculate multiple possible propagation paths from the disturbance source node to the target load area.

[0063] The propagation cost is determined based on the historical disturbance intensity, voltage fluctuation frequency and response delay coefficient between nodes;

[0064] Among all the propagation paths, the path with the lowest propagation cost is selected as the priority blocking path.

[0065] Specifically, to achieve rapid identification and isolation control of key disturbance paths, a disturbance transmission directed graph is first constructed based on the disturbance propagation graph. This directed graph uses each node in the power supply and distribution network as its vertices, and establishes directed edges based on the historical disturbance path relationships, propagation probabilities, response delays, and other information recorded in the disturbance propagation graph. Specifically, a directed edge is set between two nodes only when there is a non-zero disturbance propagation probability between them and a physical electrical connection path exists; the direction of this edge is from the upstream node to the downstream node that may be affected by the disturbance, which is used to simulate the propagation flow of the disturbance event.

[0066] Subsequently, a propagation cost is defined on each directed edge to quantify the risk and time cost of transmitting a disturbance from the source node to the downstream node. This propagation cost is composed of the following three items: (1) a historical disturbance intensity index, which reflects the cumulative propagation probability of disturbance events between two nodes; a higher value indicates more frequent propagation; (2) a voltage fluctuation frequency, which indicates the frequency of voltage anomalies at the downstream node after the disturbance; and (3) a response delay coefficient, which indicates the average delay time from the upstream disturbance to the abnormal response of the downstream node. To integrate the influence of multiple factors, a weighted cost function is used to calculate the propagation cost of each edge.

[0067] After constructing the directed graph, the minimum propagation cost path algorithm is used to search all possible paths to the target load area starting from the disturbance source node.

[0068] Further preferably, in order to cope with the actual situation of concurrent multi-path disturbances, the present invention introduces a multi-path optimization strategy, which not only evaluates the shortest propagation cost of a single path, but also evaluates the overlap rate, redundancy and coverage breadth between paths, and selects the main path with the greatest propagation risk as the priority blocking path. At the same time, in order to improve computational efficiency, optionally, based on an improved version of the Dijkstra algorithm or a path enumeration mechanism based on depth-first search, a propagation path set can be quickly identified on a directed graph and its total propagation cost calculated. Finally, the path with the lowest propagation cost is selected as the primary isolation target, and the switch disconnection instruction is issued by the edge control node, the isolation switch state is adjusted, the transmission chain of the disturbance is cut off, and the risk diffusion range is controlled to the maximum extent.

[0069] Furthermore, before the edge control node performs a power outage propagation risk level assessment on the power supply and distribution paths in the area under its jurisdiction, the method further includes:

[0070] The edge control node identifies highly sensitive nodes within its jurisdiction based on the propagation intensity index, voltage fluctuation frequency, and historical power outage response records of each node in the disturbance propagation map. These highly sensitive nodes are given higher weights in the subsequent real-time monitoring process and serve as key decision nodes in determining the power outage propagation risk level.

[0071] After determining the high-risk propagation path, when there is an abnormal disturbance signal at a highly sensitive node, the highly sensitive node is prioritized as the starting point of the high-risk propagation path to trigger the path isolation operation.

[0072] Specifically, in order to achieve a more targeted risk assessment and isolation control mechanism, before the edge control node performs the power outage propagation risk level assessment, the present invention is set to prioritize the identification of highly sensitive nodes in the area. This identification process is based on the historical characteristic indicators of each node in the disturbance propagation map, including but not limited to the node's own propagation strength index in the historical disturbance propagation, the voltage fluctuation frequency recorded by the node in the past monitoring cycle, and the node's historical power outage response records of having been the starting point or victim of abnormal propagation many times. The sensitivity score is determined based on the above indicators, and the nodes whose scores exceed the preset sensitivity threshold will be marked as highly sensitive nodes. It should be noted that the propagation strength index here is a node-level attribute used to evaluate its ability to serve as a disturbance hub, while the aforementioned propagation cost is used for path-level evaluation, and the two do not conflict.

[0073] After highly sensitive nodes are identified, the propagation risk score weighting is dynamically adjusted in subsequent real-time outage propagation risk assessments. Specifically, in the path risk scoring function, if a propagation path passes through a highly sensitive node, the path's propagation risk score is multiplied by a sensitivity weighting factor greater than 1, giving greater weight to high-risk paths that traverse highly sensitive nodes. This effectively enhances the ability to prioritize paths to critical nodes and improves the accuracy of risk perception.

[0074] Furthermore, in order to achieve rapid response in highly sensitive scenarios, when the edge control node identifies an abnormal disturbance signal (such as voltage drop, current reverse mutation, or frequent state switching, etc.) at a highly sensitive node during real-time monitoring, the path deduction and isolation pre-logic are immediately triggered. At this time, the highly sensitive node is used as a temporary disturbance source node. Based on the current disturbance propagation map, power supply path, and edge perception data, its downstream propagation path is quickly deduced, and the propagation cost and risk score of each path are calculated. If the propagation risk of any path exceeds the response threshold, the isolation control operation is immediately initiated for the shortest propagation cost path. Compared with the traditional unified scheduling response strategy, this operation can more quickly curb the initial spread of high-risk transmission and improve the timeliness and pertinence of isolation.

[0075] Furthermore, performing the path switching operation according to the preset tolerance strategy includes:

[0076] The pre-trained tolerance scoring model is called through the edge control node, and the node current fluctuation threshold, voltage anomaly duration, communication packet loss rate and load change rate obtained from the current monitoring are used as model inputs. The tolerance scoring result is output through the tolerance scoring model, and it is determined whether to trigger the backup power supply and distribution path switching operation based on the tolerance scoring result.

[0077] Furthermore, the candidate list of backup power supply and distribution paths includes priority information of each backup power supply and distribution path; the priority information is determined based on the current load state, average switching delay and voltage fluctuation value of the power supply and distribution path;

[0078] After the backup power supply and distribution path switching operation is triggered, the target backup power supply and distribution path is selected according to the priority.

[0079] Specifically, to achieve a more intelligent and robust backup power supply and distribution path switching strategy, this paper introduces a tolerance scoring model that dynamically quantifies the current state of the power supply and distribution node's ability to withstand disturbances, and uses this to determine whether to execute a path switching operation. This model, deployed on each edge control node, provides fast reasoning and local fault tolerance.

[0080] The tolerance scoring model utilizes a lightweight neural network structure. Input features include the node current fluctuation threshold (indicating the deviation between current and short-term historical current values), the duration of voltage anomalies (indicating the persistence of the disturbance), the communication packet loss rate (reflecting the stability of the monitoring link), and the load change rate (depicting the dynamic trend of the load state). Through training, the model learns the response relationship between these multi-dimensional disturbance factors and the actual triggering of switching. It then outputs a tolerance score between 0 and 1, indicating the current node's tolerance for a given disturbance scenario.

[0081] The tolerance scoring model is trained based on historical operational log data from the campus power supply and distribution system. Sample pairs are constructed. Each sample pair includes the disturbance state parameters as input features and a supervisory label indicating whether the power supply and distribution / switching path was successfully maintained under that state. During inference, the model completes scoring inference based on current real-time monitoring data. If the tolerance score falls below a preset tolerance threshold, the backup path switching logic is triggered.

[0082] After the switching logic is triggered, the most suitable switching path is selected from a preset list of candidate backup power supply and distribution paths. Each backup path in the list has a pre-calculated and dynamically maintained priority score. This priority score takes into account the following three key dimensions:

[0083] Current load status: indicates the current load level of the nodes involved in the backup path. The lower the load, the less load pressure the path will face after being enabled, and the higher the priority;

[0084] Average switching delay: The physical / logical switching time required to enable a path. The shorter the switching delay, the faster the recovery from an abnormal state and the higher the priority;

[0085] Voltage fluctuation value: The historical or current voltage stability indicator of the backup path. The smaller the fluctuation, the higher the system availability and the higher the priority.

[0086] The above tolerance scoring and path optimization operations not only improve the adaptability to minor disturbances and reduce the false trigger rate, but also have the ability to quickly restore power supply in complex disturbance scenarios.

[0087] As a preferred embodiment, in order to achieve adaptive evaluation of the disturbance status of different power supply and distribution nodes, the present invention proposes a tolerance scoring model based on the fusion of the disturbance attention mechanism and the nonlinear scoring function, which is deployed in the edge control node to evaluate the node's tolerance to the current disturbance in real time, so as to assist in determining whether it is necessary to perform a backup power supply and distribution path switching operation.

[0088] Specifically, the model input is set to the disturbance state input vector of the current power supply and distribution node Among them, I dev is the node current fluctuation threshold, reflecting the degree of deviation between the current current and the historical average current; V dur is the duration of voltage anomaly, which is used to characterize the time duration of abnormal disturbance; P loss is the packet loss rate of the communication link, reflecting the network stability of the edge node; L rate It is the load change rate, reflecting the fluctuation trend of the node power supply load.

[0089] To enhance the model's ability to understand different perturbation factors, this paper introduces a perturbation embedding mechanism and a perturbation attention scoring structure. The specific process is as follows:

[0090] In the perturbation embedding layer, the input perturbation vector is structuredly encoded through two layers of nonlinear transformation:

[0091] Z=tanh(W2·ReLU(W1·X+b1)+b2)

[0092] Among them, W1 and W2 are weight matrices; b1 and b2 are bias vectors; and Z is the perturbation embedding representation.

[0093] The attention weight of each perturbation factor is calculated through the attention mechanism:

[0094]

[0095] Among them, z i is the embedding vector of the i-th perturbation factor in the perturbation embedding; W attn is the attention transformation matrix; q is the global score vector; α i ∈[0,1] is the weight of the i-th perturbation factor.

[0096] As another preferred embodiment, the original tolerance score calculated based on the weighted calculation of each disturbance factor is T scoreIn order to enhance the model's response sensitivity to high-risk disturbance states and avoid the linear passivation effect of the scoring process, the present invention introduces an S-type nonlinear mapping function based on the original scoring to obtain the final tolerance score value T final :

[0097]

[0098] Among them, T final ∈[0,1] is the final tolerance score, the closer it is to 0, the lower the tolerance; k>0 is the curvature parameter of the mapping function, which determines the sensitivity of the score result to boundary disturbances; μ is the tolerance score critical point, corresponding to the median of the transition from tolerated to intolerable disturbances.

[0099] By setting the above function, it is ensured that the score changes slowly when the disturbance is slight, and the score drops rapidly when the disturbance approaches the critical state, which has good dynamic response performance.

[0100] Furthermore, when there is no valid path that meets the conditions in the candidate list of backup power supply and distribution paths, the method further includes:

[0101] The edge control node performs virtual subnet division and reconstruction based on the current power supply and distribution path topology to reconstruct the power supply and distribution area boundary.

[0102] Furthermore, the virtual subnet division and reconstruction operation includes:

[0103] Dynamically divide multiple power supply and distribution virtual subnets based on the current load level of each node, power flow stability and historical interconnection risk indicators, power supply and distribution path topology, and disturbance propagation map;

[0104] When it is detected that the risk level of a node in a certain subnet continues to increase, the boundary structure of the virtual subnet is automatically adjusted, and the power supply and distribution paths of the boundary nodes are reconfigured.

[0105] Furthermore, the dynamic partitioning of the virtual subnet is performed based on a partition scoring function; the scoring function calculates the association weights between nodes based on the average load value of each node, the power supply and distribution coupling strength, the historical interconnection fluctuation rate, and the consistency of power flow direction;

[0106] The edge control node clusters the node set in the power supply and distribution path diagram using a clustering algorithm based on the scoring function result, and adjusts the power supply and distribution control affiliation and the switching boundary priority strategy of each node according to the clustering result.

[0107] Specifically, to address situations where backup power supply and distribution paths cannot meet switching requirements under extreme disturbances or complex operating conditions, the present invention further introduces a virtual subnet partitioning and reconstruction mechanism to dynamically adjust the structural boundaries of the power supply and distribution area, thereby improving fault tolerance. When no paths that meet the tolerance score or priority requirements exist in the candidate list of backup power supply and distribution paths, the edge control node will automatically initiate the virtual subnet partitioning process and reconstruct the current power supply and distribution topology to build a new dispatchable area foundation.

[0108] The virtual subnet division operation is based on the node operating status and historical disturbance behavior in the current power supply and distribution path diagram, combined with the path stability, power transfer trends, and node risk linkage recorded in the disturbance propagation map, to dynamically generate multiple power supply and distribution virtual subnet units. During the specific division process, key considerations include the current load level, power flow stability, and historical interconnection volatility of each node to measure the operational coupling relationship and risk association between nodes. Nodes with high operational consistency and stable collaboration are assigned to the same virtual subnet, resulting in a compact, low-risk integrated local subsystem.

[0109] To ensure the technical rationality of the partitioning results, a multi-index fusion scoring rule is used to determine whether nodes are suitable for belonging to the same subnet unit. The scoring mainly considers the following dimensions: the smaller the difference in average load levels between node pairs, the similar operating pressure they have; the higher the coupling strength, the closer the energy transfer relationship between them; the lower the historical interconnection volatility, the greater the stability between the two in the transmission of disturbances; and the higher the consistency of power flow trends, the stronger the synergy in long-term operation. On this basis, the overall path graph is divided using a graph structure clustering algorithm built into the edge control node to form multiple new power supply and distribution virtual subnet units.

[0110] Once the division is complete, the power supply and distribution control affiliation of each subnet node will be updated, including path binding, boundary node identification, and switching strategy priorities. This ensures local autonomy of scheduling operations within the virtual subnet and cross-subnet coordination. At the same time, key boundary nodes are retained as policy buffers to handle cross-subnet path scheduling and emergency switching tasks. If a boundary node is connected to multiple subnets simultaneously, its specific path binding will be adjusted based on the current power quality, voltage stability, and node responsiveness.

[0111] During operation, if disturbances accumulate continuously within a subnet or multiple warning signals are frequently issued, the subnet is deemed to be in a locally high-risk state, triggering a dynamic adjustment process for the subnet boundary. This process, based on real-time assessments of the risk level of each node by edge control nodes, redefines high-risk areas and automatically removes risk-concentrated areas from the atomic network or introduces them into surrounding stable nodes to form a new subnet buffer structure. Simultaneously, existing boundary nodes will reconfigure their power supply and distribution paths, prioritizing the closure of transmission paths in high-risk directions and strengthening dispatch channels to low-risk areas.

[0112] The proposed virtual subnet division and boundary reconstruction mechanism is not only highly adaptive and dynamically responsive to actual disturbance evolution, but also enhances the resilience of the entire power supply and distribution network at a structural level. This allows for localized stable operation even in extreme scenarios where backup paths fail, through structural reconstruction. This provides critical support for the safe operation of power supply and distribution in large-scale industrial parks or complex building complexes.

[0113] Furthermore, the method further comprises:

[0114] After completing the path isolation and backup power supply and distribution path switching operations, record the execution status information of this round of operations and upload the execution status information to the dispatch center;

[0115] The dispatch center dynamically modifies the indicator data in the disturbance propagation map based on the accumulated feedback data from multiple rounds of execution.

[0116] Specifically, after completing path isolation and switching to alternate power supply and distribution paths, the present invention also implements a mechanism for recording path execution status and dynamically revising the graph. This mechanism tracks and records the entire process of each round of dispatch response operations, providing a detailed data foundation for subsequent graph parameter updates.

[0117] Specifically, after the path isolation and backup path switching operations are completed, the edge control node will collect and record key execution status information related to this round of operations in real time. This execution status information includes but is not limited to the following categories: first, the switching response delay, which reflects the total response time from detecting the disturbance to the full activation of the backup path; second, the path success rate, which indicates whether the backup path successfully completes the power supply task and maintains a stable working state after it is actually put into operation; third, the load recovery status, which is used to evaluate the recovery speed and quality of the load node after the switching is completed; fourth, local fluctuation change data, including the numerical changes and trend information of the current, voltage and power fluctuations before and after the switching, which is used to analyze the secondary disturbance effect caused by the switching.

[0118] This execution status information is initially collated by the edge control node and uploaded to the upper-level dispatch center via a secure communication link. After receiving multiple rounds of dispatch execution feedback from different edge regions, the dispatch center archives and aggregates it in a time series format to form a disturbance response behavior database.

[0119] Based on this database, a dynamic correction mechanism for various indicator data in the disturbance propagation map is further implemented. Specifically, the dispatch center combines the path response characteristics in historical feedback to refit the path propagation strength values ​​recorded in the disturbance propagation map, taking into account dynamic characteristics such as the fluctuation attenuation rate and load migration effect that occur in actual execution. The node response delay coefficient is corrected by iteratively updating the statistical results of the time difference between node activation, control signal reception, and load changes during multiple switching processes to make the coefficient more consistent with the actual response characteristics of the current node. The propagation risk weight is quantitatively corrected by evaluating the risk trigger frequency and failure consequences of different paths under historical disturbances.

[0120] Through the above-mentioned settings, the present invention has the ability of continuous learning and self-correction in a dynamic environment, and can improve the overall intelligence level of the power supply and distribution system without changing the hardware structure.

[0121] Finally, it should be noted that the above-mentioned embodiments include multiple parallel implementation methods of the present invention, and deleting or otherwise adjusting one or more of the implementation methods will not affect the implementation of the solution. In addition, the embodiment of the present invention discloses an intelligent operation and maintenance control method for power supply and distribution in an industrial park based on an industrial interconnection architecture, which is only a preferred embodiment of the present invention and is only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, ordinary technicians in this field should understand that it is still possible to modify the technical solutions recorded in the aforementioned embodiments, or to replace some of the technical features therein with equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for intelligent operation and maintenance of power supply and distribution in industrial parks based on industrial interconnection architecture, characterized in that: The method comprises: Based on the physical connection relationship of each power node and the power supply and distribution hierarchy information, a power supply and distribution path diagram between nodes is generated; Based on the power supply and distribution path diagram and the historical operation data and real-time status data of each node, identifying the potential power outage propagation path of each node; Based on the power supply and distribution path diagram, the potential power outage propagation path of each node and the corresponding propagation intensity index are marked as a disturbance propagation map; The disturbance propagation map is deployed to each edge control node, and the edge control node dynamically evaluates the power outage propagation risk level of the power supply and distribution path in the area under its jurisdiction based on the disturbance propagation map and the real-time monitored current, voltage and node load status data in the area under its jurisdiction. The power outage propagation risk level is used to determine whether to trigger path isolation and switching operations.

2. The method for intelligent operation and maintenance of power supply and distribution in industrial parks based on industrial interconnection architecture according to claim 1 is characterized in that: The method further comprises: Determining whether the power outage propagation risk level exceeds a preset threshold; The power supply and distribution path with a positive judgment result is regarded as a high-risk transmission path, and the path isolation operation is performed through the edge control node in the jurisdiction of the high-risk transmission path; Based on the power supply and distribution path diagram and the candidate list of backup power supply and distribution paths, the path switching operation is performed according to the preset tolerance strategy.

3. The method for intelligent operation and maintenance of power supply and distribution in industrial parks based on industrial interconnection architecture according to claim 2 is characterized in that: The path isolation operation includes: A disturbance propagation directed graph is constructed based on the disturbance propagation graph, and the minimum propagation cost path algorithm is used to calculate multiple possible propagation paths from the disturbance source node to the target load area. The propagation cost is determined based on the historical disturbance intensity, voltage fluctuation frequency and response delay coefficient between nodes; Among all the propagation paths, the path with the lowest propagation cost is selected as the priority blocking path.

4. The method for intelligent operation and maintenance of power supply and distribution in industrial parks based on industrial interconnection architecture according to claim 2 is characterized in that: Before the edge control node performs a power outage propagation risk assessment on the power supply and distribution paths in the area under its jurisdiction, the method further includes: The edge control node identifies highly sensitive nodes within its jurisdiction based on the propagation intensity index, voltage fluctuation frequency, and historical power outage response records of each node in the disturbance propagation map. These highly sensitive nodes are given higher weights in the subsequent real-time monitoring process and serve as key decision nodes in determining the power outage propagation risk level. After determining the high-risk propagation path, when there is an abnormal disturbance signal at a highly sensitive node, the highly sensitive node is prioritized as the starting point of the high-risk propagation path to trigger the path isolation operation.

5. The method for intelligent operation and maintenance of power supply and distribution in industrial parks based on industrial interconnection architecture according to claim 2 is characterized in that: The performing of the path switching operation according to the preset tolerance strategy includes: The pre-trained tolerance scoring model is called through the edge control node, and the node current fluctuation threshold, voltage anomaly duration, communication packet loss rate and load change rate obtained from the current monitoring are used as model inputs. The tolerance scoring result is output through the tolerance scoring model, and it is determined whether to trigger the backup power supply and distribution path switching operation based on the tolerance scoring result.

6. The method for intelligent operation and maintenance of power supply and distribution in industrial parks based on industrial interconnection architecture according to claim 5 is characterized in that: The candidate list of backup power supply and distribution paths includes priority information of each backup power supply and distribution path; The priority information is determined based on the current load state, average switching delay and voltage fluctuation value of the power supply and distribution path; After the backup power supply and distribution path switching operation is triggered, the target backup power supply and distribution path is selected according to the priority.

7. The method for intelligent operation and maintenance of power supply and distribution in industrial parks based on industrial interconnection architecture according to claim 2 is characterized in that: When there is no valid path that meets the conditions in the candidate list of backup power supply and distribution paths, the method further includes: The edge control node performs virtual subnet division and reconstruction based on the current power supply and distribution path topology to reconstruct the power supply and distribution area boundary.

8. The method for intelligent operation and maintenance of power supply and distribution in industrial parks based on industrial interconnection architecture according to claim 7 is characterized in that: The virtual subnet division and reconstruction operation includes: Dynamically divide multiple power supply and distribution virtual subnets based on the current load level of each node, power flow stability and historical interconnection risk indicators, power supply and distribution path topology, and disturbance propagation map; When it is detected that the risk level of a node in a certain subnet continues to increase, the boundary structure of the virtual subnet is automatically adjusted, and the power supply and distribution paths of the boundary nodes are reconfigured.

9. The method for intelligent operation and maintenance of power supply and distribution in industrial parks based on industrial interconnection architecture according to claim 8 is characterized in that: The dynamic partitioning of the virtual subnet is performed based on a partition scoring function; the scoring function calculates the association weights between nodes based on the average load value of each node, the power supply and distribution coupling strength, the historical interconnection fluctuation rate, and the consistency of power flow direction; The edge control node clusters the node set in the power supply and distribution path diagram using a clustering algorithm based on the scoring function result, and adjusts the power supply and distribution control affiliation and the switching boundary priority strategy of each node according to the clustering result.

10. The method for intelligent operation and maintenance of power supply and distribution in an industrial park based on an industrial interconnection architecture according to any one of claims 2 to 9, characterized in that: The method further comprises: After completing the path isolation and backup power supply and distribution path switching operations, record the execution status information of this round of operations and upload the execution status information to the dispatch center; The dispatch center dynamically modifies the indicator data in the disturbance propagation map based on the accumulated feedback data from multiple rounds of execution.

Images