Power distribution network operation planning method and device, computer equipment and program product

Abstract

This application relates to a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for operation planning of a distribution network. The method includes: acquiring historical operation data of the distribution network and determining the time of anomaly occurrence; collecting initial data within a preset time period before the occurrence time, preprocessing the initial data to obtain a processed power grid dataset, and constructing a directed chain graph of the distribution network; selecting the continuous behavioral sub-chain with the largest change in load response intensity value in the directed chain graph as anomaly candidate chains, determining the inducing intensity value of the anomaly candidate chains, and identifying anomaly inducing chains from among the anomaly candidate chains; matching the anomaly inducing chains with the candidate strategy behavioral chains corresponding to the current operation plan of the distribution network to obtain the behavioral chain matching degree; and based on the behavioral chain matching degree, performing avoidance correction on the current operation plan to obtain a corrected operation plan. This method can improve the safety of distribution network operation.

Description

Technical Field

[0001] This application relates to the field of smart grid planning technology, and in particular to a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for the operation planning of a distribution network. Background Technology

[0002] Power system dispatching and operation planning are crucial aspects of smart grid development and core components in the technological evolution of the energy internet, active distribution networks, and distributed dispatch systems. As the intermediate link between the power supply side and the load side, the operation of the distribution network directly impacts the stability, flexibility, and security of the entire power system. Especially in actual operation, distribution network dispatching strategies not only encompass basic tasks such as power flow allocation, voltage control, and load forecasting, but also involve multi-dimensional dynamic behaviors such as topology adjustments, distributed power source integration, and fault self-healing.

[0003] In existing distribution network operation anomaly identification and early warning mechanisms, many systems rely on single-point judgments of static indicators such as voltage, current, and load fluctuations. They primarily focus on numerical anomaly triggers at the result level, neglecting the systemic causal relationship between preceding operational behaviors, response evolution processes, and anomalies. Under this mechanism, although the system can monitor anomalies such as voltage exceeding limits or power flow reversal at a certain node, it is difficult to identify the underlying "dispatch-response" chain of causes and cannot reconstruct the operational logic that triggered the anomaly.

[0004] This fragmented identification logic is prone to missing key pre-operation nodes or weak disturbance signals in real-world scenarios such as scheduling optimization, topology adjustment, or device switching, making it impossible for the system to provide early warning and interception when the initial trigger occurs. Summary of the Invention

[0005] Therefore, it is necessary to provide a distribution network operation planning method, device, computer equipment, computer-readable storage medium, and computer program product that can improve the safety of distribution network operation planning in response to the above-mentioned technical problems.

[0006] Firstly, this application provides a method for operation planning of a power distribution network, including:

[0007] Obtain historical operating data of the power distribution network and determine the time of occurrence of the anomaly from the historical operating data;

[0008] Historical operating data within a preset time period before the occurrence time is collected as initial data, and the initial data is preprocessed to obtain the processed power grid dataset.

[0009] Based on the power grid dataset, a directed chain graph of the distribution network is constructed;

[0010] Based on the directed chain graph, determine the load response intensity value of each node in the directed chain graph;

[0011] The continuous sub-chain of behaviors with the largest change in load response intensity value in the directed chain graph is taken as an anomaly candidate chain. The cause intensity value of the anomaly candidate chain is determined, and the anomaly cause chain is determined in the anomaly candidate chain.

[0012] The abnormal cause chain is matched with the candidate strategy behavior chain corresponding to the current operation plan of the distribution network to obtain the behavior chain matching degree.

[0013] Based on the matching degree of the behavior chain, the current operation plan is modified to avoid obstacles, resulting in a modified operation plan.

[0014] In one embodiment, the method further includes:

[0015] Determine the load disturbance value of the revised operation plan;

[0016] The maximum value of the load disturbance value is compared with a preset dynamic avoidance threshold. If the maximum value is less than the dynamic avoidance threshold, the abnormal avoidance of the distribution network is determined to be successful.

[0017] If the maximum value is greater than or equal to the dynamic avoidance threshold, the abnormal avoidance of the distribution network is determined to have failed.

[0018] In one embodiment, the power grid dataset includes the control behavior sequence, topology state, and load response state of the distribution network; the step of constructing a directed chain graph of the distribution network based on the power grid dataset includes:

[0019] For each moment's control behavior sequence, determine the time delay when the degree of change of the delay function between the control behavior sequence and subsequent topology changes exceeds a preset level, and the maximum response time delay between the control behavior sequence and the node load response;

[0020] The sum of the time delay and the maximum response time delay is used as the comprehensive delay weight from the action node to the result node;

[0021] The control behavior sequence is defined as a set of behavior nodes, and the topology state and load response state are positioned as a set of result nodes.

[0022] Based on the set of behavioral nodes, the set of result nodes, and the comprehensive delay weight, a directed chain graph of the distribution network is constructed.

[0023] In one embodiment, determining the load response intensity value of each node in the directed chain graph based on the directed chain graph includes:

[0024] Extract the node sequence of the topological state of the directed chain graph, and calculate the topological offset metric of the node at each time step based on the node sequence.

[0025] Extract the load response state sequence, calculate the load change rate of all nodes based on the load response state sequence, and aggregate them to obtain the first load response intensity value of each node at each moment;

[0026] The sum of the topology offset metric and the first load response intensity value is used as the load response intensity value of each node in the directed chain graph.

[0027] In one embodiment, the step of identifying the continuous behavioral sub-chain with the largest change in load response intensity value in the directed chain graph as anomaly candidate chain, determining the inducing intensity value of the anomaly candidate chain, and identifying the anomaly inducing chain from the anomaly candidate chain includes:

[0028] Within a preset analysis interval, determine all continuous behavioral sub-chains in the directed chain graph;

[0029] The sum of the load response intensity values ​​of all behavior nodes in the continuous behavior subchain is used as the load response intensity value of the corresponding continuous behavior subchain.

[0030] The continuous behavior subchain with the largest change in load response intensity value is selected as the abnormal candidate chain;

[0031] The sum of the load response intensity values ​​of all behavioral nodes in the abnormal candidate chain is taken as the cause intensity value of the abnormal candidate chain.

[0032] Abnormal candidate chains with a trigger strength value greater than or equal to a preset trigger threshold are designated as abnormal trigger chains.

[0033] In one embodiment, the process of determining the preset trigger threshold includes:

[0034] Based on the load response intensity values ​​of all behavioral nodes within the analysis interval, the overall level and standard deviation of the disturbance intensity are determined;

[0035] The product of the standard deviation and the preset adjustment factor is used as the sensitivity index;

[0036] The number of behavioral nodes in the abnormal candidate chain that cause a change in network topology greater than a preset change level is determined, and the product of the number of behavioral nodes and a preset penalty coefficient is used as structural complexity compensation.

[0037] The sum of the overall level, the sensitivity index, and the structural complexity compensation is used as the preset trigger threshold.

[0038] Secondly, this application also provides a power distribution network operation planning device, comprising:

[0039] The determination module is used to acquire historical operating data of the power distribution network and determine the time of occurrence of the anomaly from the historical operating data;

[0040] The processing module is used to collect historical operating data within a preset time period before the occurrence time as initial data, and to preprocess the initial data to obtain the processed power grid dataset.

[0041] A construction module is used to construct a directed chain graph of the distribution network based on the power grid dataset;

[0042] The determining module is further configured to determine the load response intensity value of each node in the directed chain graph based on the directed chain graph.

[0043] The determining module is further configured to take the continuous behavior sub-chain with the largest change in load response intensity value in the directed chain graph as anomaly candidate chain, determine the cause intensity value of the anomaly candidate chain, and determine the anomaly cause chain in the anomaly candidate chain.

[0044] The matching module is used to match the abnormal cause chain with the candidate strategy behavior chain corresponding to the current operation plan of the distribution network to obtain the behavior chain matching degree.

[0045] The correction module is used to correct the current running plan based on the matching degree of the behavior chain, so as to obtain the corrected running plan.

[0046] Thirdly, this application also 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 perform the following steps:

[0047] Obtain historical operating data of the power distribution network and determine the time of occurrence of the anomaly from the historical operating data;

[0048] Historical operating data within a preset time period before the occurrence time is collected as initial data, and the initial data is preprocessed to obtain the processed power grid dataset.

[0049] Based on the power grid dataset, a directed chain graph of the distribution network is constructed;

[0050] Based on the directed chain graph, determine the load response intensity value of each node in the directed chain graph;

[0051] The continuous sub-chain of behaviors with the largest change in load response intensity value in the directed chain graph is taken as an anomaly candidate chain. The cause intensity value of the anomaly candidate chain is determined, and the anomaly cause chain is determined in the anomaly candidate chain.

[0052] The abnormal cause chain is matched with the candidate strategy behavior chain corresponding to the current operation plan of the distribution network to obtain the behavior chain matching degree.

[0053] Based on the matching degree of the behavior chain, the current operation plan is modified to avoid obstacles, resulting in a modified operation plan.

[0054] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0055] Obtain historical operating data of the power distribution network and determine the time of occurrence of the anomaly from the historical operating data;

[0056] Historical operating data within a preset time period before the occurrence time is collected as initial data, and the initial data is preprocessed to obtain the processed power grid dataset.

[0057] Based on the power grid dataset, a directed chain graph of the distribution network is constructed;

[0058] Based on the directed chain graph, determine the load response intensity value of each node in the directed chain graph;

[0059] The continuous sub-chain of behaviors with the largest change in load response intensity value in the directed chain graph is taken as an anomaly candidate chain. The cause intensity value of the anomaly candidate chain is determined, and the anomaly cause chain is determined in the anomaly candidate chain.

[0060] The abnormal cause chain is matched with the candidate strategy behavior chain corresponding to the current operation plan of the distribution network to obtain the behavior chain matching degree.

[0061] Based on the matching degree of the behavior chain, the current operation plan is modified to avoid obstacles, resulting in a modified operation plan.

[0062] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0063] Obtain historical operating data of the power distribution network and determine the time of occurrence of the anomaly from the historical operating data;

[0064] Historical operating data within a preset time period before the occurrence time is collected as initial data, and the initial data is preprocessed to obtain the processed power grid dataset.

[0065] Based on the power grid dataset, a directed chain graph of the distribution network is constructed;

[0066] Based on the directed chain graph, determine the load response intensity value of each node in the directed chain graph;

[0067] The continuous sub-chain of behaviors with the largest change in load response intensity value in the directed chain graph is taken as an anomaly candidate chain. The cause intensity value of the anomaly candidate chain is determined, and the anomaly cause chain is determined in the anomaly candidate chain.

[0068] The abnormal cause chain is matched with the candidate strategy behavior chain corresponding to the current operation plan of the distribution network to obtain the behavior chain matching degree.

[0069] Based on the matching degree of the behavior chain, the current operation plan is modified to avoid obstacles, resulting in a modified operation plan.

[0070] The aforementioned distribution network operation planning method, apparatus, computer equipment, computer-readable storage medium, and computer program product first acquire historical operation data of the distribution network and determine the occurrence time of anomalies from the historical operation data; collect historical operation data within a preset time period before the occurrence time as initial data, and preprocess the initial data to obtain a processed power grid dataset; construct a directed chain graph of the distribution network based on the power grid dataset; determine the load response intensity value of each node in the directed chain graph based on the directed chain graph; select the continuous behavioral sub-chain with the largest change in load response intensity value in the directed chain graph as anomaly candidate chain, determine the causation intensity value of the anomaly candidate chain, and determine the anomaly causation chain from the anomaly candidate chain; match the anomaly causation chain with the candidate strategy behavioral chain corresponding to the current operation plan of the distribution network to obtain the behavioral chain matching degree; and perform avoidance correction on the current operation plan based on the behavioral chain matching degree to obtain the corrected operation plan. Thus, by constructing a directed chain graph through backtracking, operational behaviors and abnormal results are dynamically linked, enhancing the interpretability of the anomaly mechanism and enabling planning adjustments to have causal chain support. By introducing load response intensity as a bridging quantity, the impact of control behaviors on the load of downstream nodes is quantified, solving the problem of the lack of continuous quantitative correlation between behavior adjustment and effect evaluation. By evaluating the matching degree between the anomaly cause chain and candidate strategies, high-risk recurrence paths can be predicted before strategy implementation, triggering chain-level avoidance correction, proactively isolating potential causes, and preventing anomaly recurrence. Attached Figure Description

[0071] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0072] Figure 1 This is an application environment diagram of the power distribution network operation planning method in one embodiment;

[0073] Figure 2 This is a flowchart illustrating the operation planning method for a power distribution network in one embodiment;

[0074] Figure 3 This is a structural block diagram of a power distribution network operation planning device in one embodiment;

[0075] Figure 4 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0076] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0077] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0078] The power distribution network operation planning method provided in this application embodiment can be applied to, for example, Figure 1In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or located on the cloud or other network servers. Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, drones, low-altitude aircraft, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, projection devices, etc. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. Head-mounted devices can be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. Server 104 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0079] In one exemplary embodiment, such as Figure 2 As shown, a method for operation planning of a distribution network is provided, which can be applied to... Figure 1 Taking terminal 102 as an example, the explanation includes the following steps 202 to 214. Wherein:

[0080] Step 202: Obtain historical operating data of the distribution network and determine the time ta when the anomaly occurred in the historical operating data.

[0081] Historical operating data refers to the operating status data of the distribution network continuously recorded by the distribution network dispatching system, monitoring system, and protection and control system during historical operation. The operating status data includes control behavior data, network topology data, and load response data.

[0082] For example, the operation status data of the distribution network during its historical operation is obtained, and the time ta of the occurrence of an anomaly in the operation status data is identified.

[0083] Step 204: Collect historical operating data within a preset time period before the occurrence time ta as initial data, and preprocess the initial data to obtain the processed power grid dataset DW.

[0084] Optionally, historical operating data within a preset time period before the occurrence time ta is collected as initial data, and the initial data is preprocessed to obtain the processed power grid dataset DW.

[0085] In one embodiment, the occurrence time ta of the anomaly is determined, and the analysis interval [ta-ΔT, ta] of a fixed window ΔT is traced back to the end point of the backtracking with the anomaly start time ta as the backtracking endpoint. Initial data is collected, and after preprocessing, the processed power grid dataset DW is obtained.

[0086] The initial data includes the control behavior sequence U(t), the topology state S(t), and the load response state P(i,t).

[0087] The control behavior sequence U(t) represents the set of control actions at time t, including power regulation, switching on / off and feeder switching.

[0088] Specifically, power regulation includes active power adjustment and reactive power adjustment. Active power adjustment is setting the output power value of distributed power sources, and reactive power adjustment is setting the reactive power through reactive power compensation devices. Switching refers to the switching operation of circuit breakers, load switches and disconnecting switches on the equipment side through remote control, changing the node connection relationship, closing and opening. Feeder switching refers to the switching of power supply paths in the feeder reconfiguration scenario.

[0089] The topology state S(t) represents the connection relationship and on / off status between all network topology nodes and branches in the power distribution network at time t during operation; the load response state P(i,t) represents the instantaneous power of the load at node i at time t.

[0090] In some embodiments, preprocessing includes removing abnormal missing points and invalid sampling points, performing dimensionless processing on data with different dimensions, unifying the dimensions of the data, aligning the data according to a unified time index, eliminating sampling frequency differences, and obtaining the processed control behavior sequence U(t), topology state S(t), and load response state P(i,t), which further constitute the processed power grid dataset DW.

[0091] Step 206: Based on the power grid dataset DW, construct the directed chain graph Gb of the distribution network.

[0092] For example, based on the power grid dataset DW, the temporal correlation between the control behavior sequence U(t), the topology state S(t), and the load response state P(i,t) is analyzed, and a directed chain graph Gb of the distribution network is constructed.

[0093] Step 208: Based on the directed chain graph Gb, determine the load response intensity value of each node in the directed chain graph Gb.

[0094] Optionally, from the directed chain graph Gb, the node sequence of the topology state S(t) is extracted to construct the time evolution sequence. With the reference structure, the topology offset metric Ds(t) at each time step is extracted. The load response state P(i,t) sequence is extracted from the power grid dataset DW. The load change rate per unit time is calculated and obtained. Then, the load change rates of all nodes are aggregated to form the first load response intensity value Dp(t) at time t. The sum of the topology offset metric Ds(t) and the first load response intensity value Dp(t) is used as the load response intensity value Fb of each node.

[0095] Step 210: The continuous behavioral sub-chain with the largest change in load response intensity value Fb in the directed chain graph Gb is taken as the abnormal candidate chain Hab. The inducing intensity value Ψab of the abnormal candidate chain is determined, and the abnormal inducing chain Cab is determined in the abnormal candidate chain.

[0096] For example, the continuous behavioral sub-chain with the largest change in load response intensity value Fb in the directed chain graph Gb is taken as the abnormal candidate chain Hab, the inducing intensity value Ψab of the abnormal candidate chain Hab is determined, and the abnormal inducing chain Cab is determined in the abnormal candidate chain Hab.

[0097] Step 212: Match the anomaly cause chain Cab with the candidate strategy behavior chain Unnew corresponding to the current operation plan of the distribution network to obtain the behavior chain matching degree Msim.

[0098] Optionally, the abnormal cause chain Cab is matched with the candidate strategy behavior chain Unnew corresponding to the current operation plan of the distribution network to obtain the behavior chain matching degree Msim.

[0099] In some embodiments, the behavior chain matching degree Msim is obtained by comparing the behavior nodes in the anomaly cause chain Cab with the candidate policy behavior chain Unew, counting the number of identical and highly similar behavior nodes, i.e. the number of intersection nodes, dividing the number of intersection nodes by the total number of behavior nodes in the anomaly cause chain Cab, and obtaining the ratio value, i.e. the behavior chain matching degree Msim.

[0100] Among them, highly similar behavior nodes are those that, although different in specific device number or node number, have the same control action type, target object, area of ​​action, or execution result; when behavior nodes have the same action type or the same effect, they are considered to be highly similar.

[0101] Step 214: Based on the behavior chain matching degree Msim, the current running plan is modified to avoid obstacles, resulting in the modified running plan Umod.

[0102] For example, when the behavior chain matching degree Msim has a potential threat of a high-risk link, chain-level matching is performed with the anomaly cause chain Cab. Based on the nodes in the anomaly cause chain Cab, a structured avoidance correction is performed on the candidate strategy behavior chain Unew to obtain the corrected running plan Umod.

[0103] In some embodiments, the candidate policy behavior chain Unew, which is to be executed in the current cycle, is compared and analyzed with the stored anomaly cause chain Cab to determine which nodes in the candidate policy behavior chain have overlapping behaviors with the anomaly cause chain Cab. Then, the rule-driven chain avoidance function is called to generate the avoidance-corrected behavior chain, i.e., the corrected execution plan Umod.

[0104] In the above-mentioned distribution network operation planning method, historical operation data of the distribution network is acquired, and the occurrence time of an anomaly is determined from the historical operation data; historical operation data within a preset period before the occurrence time is collected as initial data, and the initial data is preprocessed to obtain the processed power grid dataset DW; based on the power grid dataset DW, a directed chain graph Gb of the distribution network is constructed; based on the directed chain graph Gb, the load response intensity value Fb of each node in the directed chain graph Gb is determined; the continuous behavioral sub-chain with the largest change in load response intensity value Fb in the directed chain graph Gb is taken as the anomaly candidate chain Hab, the cause intensity value Ψab of the anomaly candidate chain Hab is determined, and the anomaly cause chain Cab is determined in the anomaly candidate chain Hab; the anomaly cause chain Cab is matched with the candidate strategy behavior chain Unew corresponding to the current operation plan of the distribution network to obtain the behavior chain matching degree Msim; based on the behavior chain matching degree Msim, the current operation plan is modified to avoidance, and the modified operation plan Umod is obtained. Thus, by constructing a directed chain graph Gb through backtracking, operational behaviors and abnormal results are dynamically linked, enhancing the interpretability of the anomaly mechanism and enabling planning adjustments to have causal chain support. By introducing load response intensity as a bridging quantity, the impact of control behaviors on the load of downstream nodes is quantified, solving the problem of the lack of continuous quantitative correlation between behavior adjustment and effect evaluation. By evaluating the matching degree between the anomaly trigger chain Cab and candidate strategies, high-risk recurrence paths can be predicted before strategy implementation, triggering chain-level avoidance correction, actively isolating potential triggers, and blocking the recurrence of anomalies.

[0105] In an exemplary embodiment, the operation planning method for a distribution network further includes: determining the load disturbance value nFb of the modified operation plan Umod; comparing the maximum value of the load disturbance value nFb with a preset dynamic avoidance threshold; if the maximum value is less than the dynamic avoidance threshold, determining that the distribution network has successfully avoided an anomaly; if the maximum value is greater than or equal to the dynamic avoidance threshold, determining that the distribution network has failed to avoid an anomaly.

[0106] In actual implementation, the revised operation plan Umod is re-analyzed, and the load disturbance value nFb of the behavior chain in the revised operation plan Umod is extracted. The maximum value of the load disturbance value nFb is compared with the preset dynamic avoidance threshold. If the maximum value is less than the dynamic avoidance threshold, the abnormal avoidance of the distribution network is determined to be successful; if the maximum value is greater than or equal to the dynamic avoidance threshold, the abnormal avoidance of the distribution network is determined to be unsuccessful.

[0107] In some embodiments, the load response state P(i,t) of each scheduling node i in the behavior chain at different time points t is obtained; the load disturbance value nFb is constructed by performing square accumulation on the power change rate of each scheduling node over time.

[0108] Among them, the dispatch node is the basic unit in the operation of the distribution network that can reflect the system status, execute control commands and generate load response.

[0109] In some embodiments, when the avoidance strategy is valid under the current operating conditions, the obtained modified operating plan Umod is bound to the corresponding original anomaly trigger chain Cab to form a safe avoidance mapping relationship and stored in the safe chain memory. When the avoidance strategy is not valid, the high-disturbance behavior chain actually formed during this execution is re-extracted and added to the anomaly trigger chain database as a candidate for a new anomaly trigger chain.

[0110] In the above embodiments, by re-analyzing the load disturbance value nFb of the corrected behavior chain Umod and comparing it with the dynamic threshold, a closed-loop quantitative verification of the avoidance effect is achieved, significantly improving the credibility and adaptability of the unconventional avoidance strategy. This mechanism can objectively evaluate the disturbance intensity caused to the system by the corrective operation after actual execution. By comparing the maximum disturbance value with the dynamic avoidance threshold, it accurately determines whether the avoidance is successful, avoiding the blindness of traditional methods that rely solely on experience or static indicators.

[0111] In an exemplary embodiment, the power grid dataset DW includes a control behavior sequence U(t), a topology state S(t), and a load response state P(i,t) of the distribution network. Based on the power grid dataset DW, a directed chain graph Gb of the distribution network is constructed, including: for each moment of the control behavior sequence U(t), determining the time delay between the change in the delay function of the subsequent topology changes and the maximum response time delay between the control behavior sequence U(t) and the load response of the nodes; using the sum of the time delay and the maximum response time delay as the comprehensive delay weight from the behavior node to the result node; defining the control behavior sequence U(t) as a set of behavior nodes, and positioning the topology state S(t) and the load response state P(i,t) as a set of result nodes; and constructing the directed chain graph Gb of the distribution network based on the set of behavior nodes, the set of result nodes, and the comprehensive delay weight.

[0112] In actual implementation, for each control behavior sequence U(t) at any given time, within a fixed period, the delay relationship between behavior and response is constructed, including the time delay △Ts(t) when the degree of change of the delay function between subsequent topology changes exceeds the preset level, and the maximum response time delay △Tp(t) between the node load response and the control behavior sequence U(t).

[0113] In some embodiments, the time delay ΔTs(t) is obtained by: starting from the moment when the action of the control behavior sequence U(t) occurs, searching backward for a fixed period of time (t+tk), and reading the topological state S(t+tk) at time point (t+tk) and comparing it with the topological state S(t).

[0114] Based on the L1 norm structural difference measurement method, the difference between two structural states is measured. If the difference between the structural states is greater than the preset minimum significant change threshold at a future time point, the time difference is identified as the time delay △Ts(t) of a significant change (the degree of change of the delay function exceeds the preset level).

[0115] In some embodiments, for the control behavior sequence U(t) and the load response state P(i,t), the time delay function from the behavior to the significant load response is defined as the maximum response time delay ΔTp(t). The specific method for obtaining this function is as follows: Starting from the time when the control behavior sequence U(t) occurs, the load change trend of the node is continuously observed towards future time points, and the rate of change of the load response curve of node i in the subsequent time period is calculated, i.e., the derivative of the load response state P(i,t) is taken; when the rate of change of the load response curve of node i reaches its maximum value at the future time point (t+tk), the time delay between the occurrence of the behavior and the strongest load response of the node is obtained according to the corresponding time point, i.e., the maximum response time delay ΔTp(t).

[0116] In some embodiments, the sum of the time delay ΔTs(t) and the maximum response time delay ΔTp(t) is used as the comprehensive delay weight T(tk,i) from the action node to the result node.

[0117] In some embodiments, the control behavior sequence U(t) is defined as the behavior node set Vu, and the topology state S(t) and load response state P(i,t) are located as the result node set Vr; specifically, the behavior node set Vu={U(t1), U(t2), ..., U(tn)}; the result node set Vr={S(tj), P(i,tj)|tj∈[ta-ΔT,ta]}; tj represents the time point.

[0118] In some embodiments, when there is a significant change relationship between the control behavior sequence U(t) and the topology state S(t) or between the control behavior sequence U(t) and the load response state P(i,t), a directed edge set E→{S(t+tk), P(i,t+tk)} and edge weight wk are established; the edge weight wk is obtained by the reciprocal of the comprehensive delay weight T(tk,i). Further, a directed chain graph Gb=(Vu∪Vr,E,W) of the distribution network is constructed, where W represents the edge weight set of edge weight wk.

[0119] In the above embodiments, by calculating the time delay between control behavior and topology changes and load response, and constructing a comprehensive delay weight, the temporal relationship and intensity of the impact of each scheduling operation on the system state can be accurately quantified. Presenting the directed chain graph Gb as behavior nodes, result nodes, and their edge weights makes the causal path between operation behavior and system response clear and traceable, overcoming the problem of separated behavior and result analysis in traditional methods.

[0120] In an exemplary embodiment, determining the load response intensity value Fb of each node in the directed chain graph Gb includes: extracting the node sequence of the topology state S(t) in the directed chain graph Gb; calculating the topology offset metric Ds(t) of the node at each time step based on the node sequence; extracting the load response state sequence; calculating and aggregating the load change rate of all nodes based on the load response state sequence to obtain the first load response intensity value Dp(t) of the node at each time step; and using the sum of the topology offset metric Ds(t) and the first load response intensity value Dp(t) as the load response intensity value Fb of each node in the directed chain graph Gb.

[0121] In practical implementation, the node sequence of the topology state S(t) in the directed chain graph Gb is extracted. Based on the node sequence, the topology offset metric Ds(t) of the node at each time step is calculated. The load response state sequence is extracted. Based on the load response state sequence, the load change rate of all nodes is calculated and aggregated to obtain the first load response intensity value Dp(t) of the node at each time step. The sum of the topology offset metric Ds(t) and the first load response intensity value Dp(t) is used as the load response intensity value Fb of each node in the directed chain graph Gb.

[0122] In some embodiments, the topology offset metric Ds(t) is obtained as follows: within the pre-anomaly analysis interval [ta-ΔT, ta], the starting topology state S(ta-ΔT) of the pre-anomaly analysis interval is selected as the initial structure state, and the topology state S(t) is a node attribute in the directed chain graph Gb; the degree of deviation between the current structure and the reference structure is measured by the L1 norm, and the topology offset metric Ds(t) = ||S(t)-S(ta-ΔT)||1 ​​is obtained.

[0123] In some embodiments, the first load response intensity value Dp(t) is obtained by: traversing all the key nodes that need to be monitored, a total of N nodes, calculating the load change rate at time t for each node i, squaring the instantaneous change rate of each node, and summing the squared values ​​of the instantaneous change rates of all nodes to obtain the first load response intensity value Dp(t).

[0124] Among them, key nodes are those that are representative of system load disturbance identification and abnormal evolution analysis within the pre-abnormal analysis period; key nodes include at least one of the following: nodes directly affected by control behavior, nodes with significant load changes, nodes located at feeder switching or switch operation positions, nodes located at network partition boundaries, and nodes that have appeared multiple times in historical abnormal events.

[0125] In some embodiments, the load change rate is calculated as follows: The load power values ​​of node i at two adjacent sampling times are read from the power grid dataset DW, namely the node load power value at the current sampling time t and the node load power value at the previous sampling time t-Δt, where Δt is the data sampling time interval; the difference between the node load power value at the current sampling time and the node load power value at the previous sampling time is calculated to obtain the load change of the node within that time interval; then, the ratio of this load change to the sampling time interval Δt is calculated to obtain the load change rate of node i at time t.

[0126] In the above embodiments, by constructing a significantly varying time delay ΔTs(t) and a maximum response time delay ΔTp(t), causal time offset analysis between control behavior and system response is achieved. This mechanism overcomes the problems of "unclear matching between control behavior and response state" and "lack of time-series causal inference support" in the prior art, and can form source-tracing support logic for abnormal causes before grid operation disturbances.

[0127] In an exemplary embodiment, the continuous behavioral sub-chain with the largest change in load response intensity value Fb in the directed chain graph Gb is designated as an anomaly candidate chain Hab. The causative intensity value Ψab of the anomaly candidate chain Hab is determined, and an anomaly causative chain Cat is identified within the anomaly candidate chain Hab. This includes: determining all continuous behavioral sub-chains in the directed chain graph Gb within a preset analysis interval; using the sum of the load response intensity values ​​Fb of all behavioral nodes in the continuous behavioral sub-chain as the load response intensity value Fb of the corresponding continuous behavioral sub-chain; designating the continuous behavioral sub-chain with the largest change in load response intensity value Fb as an anomaly candidate chain Hab; using the sum of the load response intensity values ​​Fb of all behavioral nodes in the anomaly candidate chain Hab as the causative intensity value Ψab of the anomaly candidate chain Hab; and designating the anomaly candidate chain Hab whose causative intensity value Ψab is greater than or equal to a preset causative threshold as the anomaly causative chain Cat.

[0128] In practice, in the directed chain graph Gb, the load response loudness value Fb of each behavior node is mapped using behavior nodes as indices. Within the entire pre-anomaly analysis interval [ta-ΔT, ta], the behavior node sequence is scanned with a fixed sliding window to find the continuous behavior sub-chain with the largest cumulative load response intensity value Fb, and is denoted as the anomaly candidate chain Hab.

[0129] In some embodiments, a sliding window of length L is used to slide along the sequence of behavior nodes in chronological order throughout the entire behavior chain; for each behavior chain segment covered by the sliding window, the corresponding load response intensity value Fb is extracted sequentially; the sum of the load response intensity values ​​Fb of all behavior nodes on the behavior chain segment is calculated to obtain the total disturbance value; finally, all sliding window segments are compared, and the continuous sub-chain with the largest total disturbance value is found as the anomalous candidate chain Hab.

[0130] In some embodiments, all behavioral nodes in the abnormal candidate chain Hab are traversed; for each behavioral node, the corresponding load response intensity value Fb is found, and all load response intensity values ​​Fb are summed to obtain the total load response, which is the cause chain intensity value Ψab.

[0131] In some embodiments, the inducing strength value Ψab of the abnormal candidate chain Hab is compared with a preset inducing threshold. If the inducing strength value Ψab is greater than the preset inducing threshold, the abnormal candidate chain Hab is determined to be the abnormal inducing chain Cab.

[0132] In the above embodiments, the entire sequence of behavioral nodes is scanned by a sliding window, and the continuous sub-chain with the strongest perturbation effect is actively searched in the pre-abnormal analysis interval. This effectively overcomes the problem in the prior art that it is impossible to extract key abnormal paths in multiple behavioral interferences, and enhances the system's extraction accuracy for complex abnormal precursors.

[0133] In an exemplary embodiment, the process of determining the preset trigger threshold includes: determining the overall level and standard deviation of the disturbance intensity based on the load response intensity value Fb of all behavioral nodes within the analysis interval; using the product of the standard deviation and a preset adjustment factor as the sensitivity index; determining the number of behavioral nodes in the abnormal candidate chain Hab that cause a change in network topology greater than a preset change level, and using the product of the number of behavioral nodes and a preset penalty coefficient as the structural complexity compensation; and using the sum of the overall level, the sensitivity index, and the structural complexity compensation as the preset trigger threshold.

[0134] In practice, within the analysis time interval, the load response intensity value Fb caused by all behavioral nodes is averaged to obtain the overall level of disturbance intensity, and the standard deviation is calculated. The standard deviation is multiplied by the adjustment factor to obtain the sensitivity index. The number of behavioral nodes in the abnormal candidate chain Hab that cause a change in network topology greater than the preset change level is determined, and the product of the number of behavioral nodes and the preset penalty coefficient is used as the structural complexity compensation. The sum of the overall level, the sensitivity index, and the structural complexity compensation is used as the preset trigger threshold.

[0135] In some embodiments, a behavior node that causes a significant change in the network structure refers to a behavior node that, after performing a control action, causes a change in the topology connection relationship of the distribution network; the change in topology connection relationship includes changes in branch opening status, switching of power supply paths, changes in network partition connectivity, or changes in the status of tie switches.

[0136] In the above embodiments, by introducing structural jump counting and sensitivity index compensation into the determination of the causal chain strength value, this method not only considers the intensity of the load response but also the network structure jump effect it causes, forming a more comprehensive logic for judging the value of candidate abnormal chains and filling the analytical blind spot of only looking at the response intensity and ignoring the structural influence. The causal threshold is not a preset constant but is dynamically generated by combining the current global disturbance average level, standard deviation fluctuation characteristics, and structural variation factors, making the abnormal chain identification more environmentally adaptable and solving the problem that fixed thresholds are not applicable to different working conditions.

[0137] To illustrate the distribution network operation planning method in this application in detail, an embodiment is provided below. For example, this application describes the distribution network operation planning method in a specific scenario.

[0138] First, acquire the operating status data of the distribution network during its historical operation and identify the occurrence time ta of the anomaly in the operating status data. After determining the occurrence time ta of the anomaly, use the anomaly start time ta as the backtracking endpoint and backtrack forward for a fixed window ΔT analysis interval [ta-ΔT, ta] to collect initial data. After preprocessing, the processed power grid dataset DW is obtained.

[0139] Based on the power grid dataset DW, this study analyzes the temporal correlation between the control behavior sequence U(t), topology state S(t), and load response state P(i,t), and constructs a directed chain graph Gb of the distribution network. From the directed chain graph Gb, the node sequence of the topology state S(t) is extracted to construct a time evolution sequence. Using the baseline structure as a reference, the topology offset metric Ds(t) is extracted at each moment. The load response state P(i,t) sequence is extracted from the power grid dataset DW, and the load change rate per unit time is calculated. The load change rates of all nodes are then aggregated to form the first load response intensity value Dp(t) at time t. The sum of the topology offset metric Ds(t) and the first load response intensity value Dp(t) is used as the load response intensity value Fb for each node.

[0140] The continuous behavioral sub-chain with the largest change in load response intensity value Fb in the directed chain graph Gb is taken as the anomaly candidate chain Hab. The inducing intensity value Ψab of the anomaly candidate chain Hab is determined, and the anomaly inducing chain Cab is determined in the anomaly candidate chain Hab. The anomaly inducing chain Cab is matched with the candidate strategy behavioral chain Unew corresponding to the current operation plan of the distribution network to obtain the behavioral chain matching degree Msim.

[0141] When a high-risk link is found in the behavior chain matching degree Msim, chain-level matching is performed with the abnormal cause chain Cab. Based on the nodes in the abnormal cause chain Cab, a structured avoidance correction is performed on the candidate strategy behavior chain Unnew to obtain the corrected running plan Umod.

[0142] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0143] Based on the same inventive concept, this application also provides a distribution network operation planning device for implementing the above-mentioned distribution network operation planning method. The solution provided by this device is similar to the implementation scheme described in the above method. Therefore, the specific limitations of one or more distribution network operation planning device embodiments provided below can be found in the limitations of the distribution network operation planning method described above, and will not be repeated here.

[0144] In one exemplary embodiment, such as Figure 3 As shown, a power distribution network operation planning device is provided, comprising: a determination module 301, a processing module 302, a construction module 303, a matching module 304, and a correction module 305, wherein:

[0145] The determination module is used to acquire historical operating data of the power distribution network and determine the time when the anomaly occurred from the historical operating data.

[0146] The processing module is used to collect historical operating data within a preset time period before the occurrence time as initial data, and to preprocess the initial data to obtain the processed power grid dataset.

[0147] A construction module is used to construct a directed chain graph of the distribution network based on the power grid dataset;

[0148] The determining module is further configured to determine the load response intensity value of each node in the directed chain graph based on the directed chain graph.

[0149] The determining module is further configured to take the continuous behavioral sub-chain with the largest change in load response intensity value in the directed chain graph as anomaly candidate chain, determine the cause intensity value of the anomaly candidate chain, and determine the anomaly cause chain in the anomaly candidate chain.

[0150] The matching module is used to match the abnormal cause chain with the candidate strategy behavior chain corresponding to the current operation plan of the distribution network to obtain the behavior chain matching degree.

[0151] The correction module is used to correct the current running plan based on the matching degree of the behavior chain, so as to obtain the corrected running plan.

[0152] In some exemplary embodiments, the above-described apparatus further includes a determination module, configured to:

[0153] Determine the load disturbance value of the revised operation plan;

[0154] The maximum value of the load disturbance is compared with the preset dynamic avoidance threshold. If the maximum value is less than the dynamic avoidance threshold, the abnormal avoidance of the distribution network is determined to be successful.

[0155] If the maximum value is greater than or equal to the dynamic avoidance threshold, the abnormal avoidance of the distribution network is determined to have failed.

[0156] In some exemplary embodiments, the above-described building module is further configured to:

[0157] For each moment's control behavior sequence, determine the time delay when the degree of change of the delay function between the control behavior sequence and subsequent topology changes exceeds a preset level, and the maximum response time delay between the control behavior sequence and the node load response;

[0158] The sum of the time delay and the maximum response time delay is used as the comprehensive delay weight from the action node to the result node;

[0159] The control behavior sequence is defined as a set of behavior nodes, and the topology state and load response state are positioned as a set of result nodes.

[0160] A directed chain graph of the distribution network is constructed based on the set of behavioral nodes, the set of result nodes, and the comprehensive delay weight.

[0161] In some exemplary embodiments, the determining module described above is further configured to:

[0162] Extract the node sequence of the topological state in the directed chain graph, and calculate the topological offset metric of the node at each time step based on the node sequence.

[0163] Extract the load response state sequence, calculate the load change rate of all nodes based on the load response state sequence, and aggregate them to obtain the first load response intensity value of each node at each time step.

[0164] The sum of the topology offset metric and the first load response intensity value is used as the load response intensity value of each node in the directed chain graph.

[0165] In some exemplary embodiments, the determining module described above is further configured to:

[0166] Within the preset analysis interval, determine all continuous behavioral sub-chains in the directed chain graph;

[0167] The sum of the load response intensity values ​​of all behavior nodes in the continuous behavior subchain is used as the load response intensity value of the corresponding continuous behavior subchain.

[0168] The continuous behavior subchain with the largest change in load response intensity value is selected as the abnormal candidate chain;

[0169] The sum of the load response intensity values ​​of all behavioral nodes in the abnormal candidate chain is used as the cause intensity value of the abnormal candidate chain.

[0170] Abnormal candidate chains with a trigger strength value greater than or equal to a preset trigger threshold are designated as abnormal trigger chains.

[0171] In some exemplary embodiments, the determining module described above is further configured to:

[0172] Based on the load response intensity values ​​of all behavioral nodes within the analysis interval, the overall level and standard deviation of the disturbance intensity are determined;

[0173] The product of the standard deviation and the preset adjustment factor is used as the sensitivity index;

[0174] The number of behavioral nodes in the abnormal candidate chain that cause a change in network topology greater than a preset change is determined, and the product of the number of behavioral nodes and the preset penalty coefficient is used as compensation for structural complexity.

[0175] The sum of the overall level, sensitivity index, and structural complexity compensation is used as the preset trigger threshold.

[0176] Each module in the aforementioned power distribution network operation planning device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the computer device's memory as software, so that the processor can call and execute the corresponding operations of each module.

[0177] In one exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 4As shown, the computer device includes a processor, memory, input / output interfaces, a communication interface, a display unit, and an input device. The processor, memory, and input / output interfaces are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The input / output interfaces are used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When the computer program is executed by the processor, it implements a power distribution network operation planning method.

[0178] The display unit of this computer device is used to form a visually visible image and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of this computer device can be a touch layer covering the display screen, or buttons, a trackball, or a touchpad set on the casing of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0179] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0180] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0181] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0182] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0183] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0184] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0185] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0186] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for operation planning of a power distribution network, characterized in that, The method includes: Obtain historical operating data of the power distribution network and determine the time of occurrence of the anomaly from the historical operating data; Historical operating data within a preset time period before the occurrence time is collected as initial data, and the initial data is preprocessed to obtain the processed power grid dataset. Based on the power grid dataset, a directed chain graph of the distribution network is constructed; Based on the directed chain graph, determine the load response intensity value of each node in the directed chain graph; Within a preset analysis interval, determine all continuous behavioral sub-chains in the directed chain graph; The sum of the load response intensity values ​​of all behavior nodes in the continuous behavior subchain is used as the load response intensity value of the corresponding continuous behavior subchain. The continuous sub-chain of behaviors with the largest change in load response intensity value is selected as the candidate chain for anomalies; The sum of the load response intensity values ​​of all behavioral nodes in the abnormal candidate chain is taken as the cause intensity value of the abnormal candidate chain. Abnormal candidate chains with inducing intensity values ​​greater than or equal to a preset inducing threshold are designated as abnormal inducing chains. The abnormal cause chain is matched with the candidate strategy behavior chain corresponding to the current operation plan of the distribution network to obtain the behavior chain matching degree. Based on the matching degree of the behavior chain, the current operation plan is modified to avoid obstacles, resulting in a modified operation plan.

2. The method of claim 1, wherein, The method further includes: Determine the load disturbance value of the revised operation plan; The maximum value of the load disturbance value is compared with a preset dynamic avoidance threshold. If the maximum value is less than the dynamic avoidance threshold, the abnormal avoidance of the distribution network is determined to be successful. If the maximum value is greater than or equal to the dynamic avoidance threshold, the abnormal avoidance of the distribution network is determined to have failed.

3. The method of claim 1, wherein, The power grid dataset includes the control behavior sequence, topology state, and load response state of the distribution network; the construction of a directed chain graph of the distribution network based on the power grid dataset includes: For each moment's control behavior sequence, determine the time delay when the degree of change of the delay function between the control behavior sequence and subsequent topology changes exceeds a preset level, and the maximum response time delay between the control behavior sequence and the node load response; The sum of the time delay and the maximum response time delay is used as the comprehensive delay weight from the action node to the result node; The control behavior sequence is defined as a set of behavior nodes, and the topology state and load response state are defined as a set of result nodes. Based on the set of behavioral nodes, the set of result nodes, and the comprehensive delay weight, a directed chain graph of the distribution network is constructed.

4. The method of claim 3, wherein, The determination of the load response intensity value of each node in the directed chain graph, based on the directed chain graph, includes: Extract the node sequence of the topological state of the directed chain graph, and calculate the topological offset metric of the node at each time step based on the node sequence. Extract the load response state sequence, and based on the load response state sequence, calculate the load change rate of all nodes and aggregate them to obtain the first load response intensity value of each node at each moment. The sum of the topology offset metric and the first load response intensity value is used as the load response intensity value of each node in the directed chain graph.

5. The method of claim 1, wherein, The process of determining the preset trigger threshold includes: Based on the load response intensity values ​​of all behavioral nodes within the analysis interval, the overall level and standard deviation of the disturbance intensity are determined; The product of the standard deviation and the preset adjustment factor is used as the sensitivity index; The number of behavioral nodes in the abnormal candidate chain that cause a change in network topology greater than a preset change level is determined, and the product of the number of behavioral nodes and a preset penalty coefficient is used as structural complexity compensation. The sum of the overall level, the sensitivity index, and the structural complexity compensation is used as the preset trigger threshold.

6. An operation planning device of a power distribution network, characterized by comprising: The device includes: The determination module is used to acquire historical operating data of the power distribution network and determine the time of occurrence of the anomaly from the historical operating data; The processing module is used to collect historical operating data within a preset time period before the occurrence time as initial data, and to preprocess the initial data to obtain the processed power grid dataset. A construction module is used to construct a directed chain graph of the distribution network based on the power grid dataset; The determining module is further configured to determine the load response intensity value of each node in the directed chain graph based on the directed chain graph. The determining module is further configured to: determine all continuous behavioral sub-chains in the directed chain graph within a preset analysis interval; take the sum of the load response intensity values ​​of all behavioral nodes in the continuous behavioral sub-chain as the load response intensity value of the corresponding continuous behavioral sub-chain; take the continuous behavioral sub-chain with the largest change in load response intensity value as an abnormal candidate chain; take the sum of the load response intensity values ​​of all behavioral nodes in the abnormal candidate chain as the cause intensity value of the abnormal candidate chain; and take the abnormal candidate chain with a cause intensity value greater than or equal to a preset cause threshold as an abnormal cause chain. The matching module is used to match the abnormal cause chain with the candidate strategy behavior chain corresponding to the current operation plan of the distribution network to obtain the behavior chain matching degree. The correction module is used to correct the current running plan based on the matching degree of the behavior chain, so as to obtain the corrected running plan.

7. 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 method according to any one of claims 1 to 5.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

9. A computer program product comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images