Power distribution network multi-time scale coordinated operation deduction method, system and device and storage medium

Abstract

The invention discloses a power distribution network multi-time scale coordinated operation deduction method, system and device and a storage medium, and relates to the technical field of power distribution networks, and the method comprises the steps: constructing an event propagation graph and a propagation path set based on the real-time operation data of a power distribution network, and carrying out the event-based deduction of the operation state of the power distribution network; calling an operation strategy library, judging whether a structure adjustment action needs to be executed or not, generating an adjusted physical topological graph and an event propagation path set, extending to multi-scene comparison deduction, and calculating scene difference indexes; if the scene difference index is greater than a preset threshold value, outputting a regulation and control suggestion sequence, otherwise, outputting a unified control strategy; according to the method, the operation state deduction of the power distribution network has logical coherence and dynamic feedback, and fault and state response can be accurately simulated at multiple time scales. A topology and state linkage closed loop is realized, and a practical regulation and control strategy is output. The method can improve the capabilities of intelligent scheduling, emergency processing and the like, and provides powerful guarantee for operation analysis of a complex power distribution network.

Description

Technical Field

[0001] This invention relates to the field of power distribution network technology, and in particular to a method, system, equipment, and storage medium for multi-timescale coordinated operation simulation of power distribution networks. Background Technology

[0002] With the large-scale integration of new energy sources, increased load fluctuations in the power system, and the increasing diversification of electricity consumption behaviors, the operating status of distribution networks is increasingly exhibiting comprehensive characteristics of "alternating multiple scenarios," "dynamic changes in operating status," and "frequent adjustments to topology." As a crucial intermediary link connecting power sources and users, the operating status of the distribution network not only directly affects the security and reliability of power supply but also plays a decisive role in upper-level dispatch optimization and lower-level load management. Against this backdrop, accurate, efficient, and dynamic simulation of distribution network operation, especially considering the evolution of normal operation and fault scenarios simultaneously across multiple time scales, has become a critical technical problem urgently needing to be solved in current intelligent dispatch and simulation systems for distribution networks.

[0003] Currently, mainstream distribution network operation analysis methods can be mainly divided into two categories: one is the real-time state simulation method based on power system simulation, which derives the operating state of each node by performing static or quasi-static calculations on state variables such as load and power flow during the operating period. Although this type of method has a certain degree of accuracy and physical consistency under normal operating scenarios, it lacks the ability to model the state evolution path, especially when facing fault disturbances, equipment failures, or dynamic changes in topology, and cannot effectively simulate the system's cascading response process and structural adaptive behavior; the other type of method focuses on optimization or control, using strategies such as mixed integer programming and reinforcement learning to find the optimal topology configuration or scheduling action at a certain moment. However, these methods usually optimize within a single time scale, and the modeling of topology changes is often fragmented and externally forced, making it difficult to form a continuous evolution closed loop. In addition, existing technologies generally regard "operating state deduction" and "topology reconstruction" as two independent tasks, lacking a technical system that can unify the modeling of the causal relationship between state evolution and structural changes, resulting in difficulty in making a fast and accurate response to the dual linkage of state and structure in actual scheduling. More seriously, in actual operation, distribution networks often exhibit multi-timescale characteristics: in the short term, the occurrence and propagation of electrical faults are highly transient and chain-triggered; while in the medium to long term, system load evolution, distributed energy access, and operational mode adjustments cause continuous changes and evolutions in the topology. Existing methods rarely cover the operational evolution process at the minute, hour, or even day level simultaneously, and cannot support key applications such as emergency simulation, operational drills, or forward-looking operation and maintenance decisions in dispatching auxiliary systems. Therefore, there is an urgent need for a new distribution network operation simulation method and system framework with "state-structure linkage modeling," "normal + fault scenario integrated simulation," and "cross-timescale continuous extrapolation capability," to fundamentally improve the systematicness, accuracy, and predictive ability of operational analysis. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by this invention is: how to provide a new distribution network operation simulation method with "state-structure linkage modeling", "normal + fault scenario integrated simulation" and "cross-time scale continuous simulation capability".

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for simulating the coordinated operation of a distribution network across multiple time scales, comprising: Based on real-time operation data of the power distribution network, an event propagation graph and a set of propagation paths are constructed. Based on the event propagation graph and propagation path set, the operation status of the distribution network is extrapolated based on events, and the evolution sequence of the operation status of the distribution network over time and the set of activated events in each time step are output. The operation strategy library is invoked to determine whether structural adjustment actions need to be performed based on the evolution sequence and the set of activated events, and to generate the adjusted physical topology and event propagation path set. The feedback process of the adjusted physical topology map and event propagation path set is extended to multi-scenario comparative simulation, and scenario difference indicators are calculated; If the scenario difference index is greater than the preset threshold, a regulation suggestion sequence will be output; otherwise, a unified control strategy will be output.

[0007] As a preferred scheme for the multi-timescale coordinated operation simulation method of distribution networks, the following is provided: The construction of the event propagation graph and propagation path set based on real-time operation data of the distribution network includes: Based on real-time operation data of the power distribution network, a physical topology diagram, an initial state vector, and an initial set of abnormal events are constructed. Based on the physical topology graph, initial state vector, and initial abnormal event set, a propagation rule generation function is designed to determine whether an event can propagate between nodes. The physical topology graph, initial state vector, and initial abnormal event set are input into the propagation weight function, which outputs the propagation weight value. Edges with propagation weight values ​​greater than a threshold are included in the propagation edge set to form the event propagation graph. Based on the event propagation graph, a propagation path set is constructed, with each path recording the event sequence, cumulative delay time, and total propagation weight.

[0008] The beneficial effects of this preferred technical solution are as follows: By constructing a physical topology graph, initial state vector, and initial abnormal event set based on real-time operation data of the distribution network, the current actual operating status of the distribution network can be accurately reflected. The designed propagation rule generation function can effectively determine the propagation probability of events between nodes. Combined with the propagation weight function, edges with propagation weight values ​​greater than a threshold are selected to form an event propagation graph, ensuring that the event propagation graph focuses on connections with actual propagation significance. Based on this, a propagation path set is constructed, recording the event sequence, cumulative delay time, and total propagation weight in detail. This provides comprehensive and accurate basic information for subsequent event-based inferences, helping to more accurately simulate the propagation process of events in the distribution network.

[0009] As a preferred scheme for the multi-timescale coordinated operation simulation method of distribution networks, the following is provided: If the scenario difference index is greater than a preset threshold, a regulation suggestion sequence is output; otherwise, a unified control strategy is output, including: The calculated scenario difference index is compared with a preset threshold. If the scenario difference index exceeds the preset threshold, the system determines that the fault has a significant impact and outputs a sequence of control suggestions for the fault scenario. If the scenario difference index does not exceed the preset threshold, the output includes the difference assessment index, control suggestions under the fault scenario, and unified control strategy for different scenario simulation trajectories used for drills and playback.

[0010] The beneficial effects of this preferred technical solution are as follows: It compares the calculated scenario difference index with a preset threshold and outputs different strategies based on the comparison results. When the scenario difference index exceeds the preset threshold, it determines that the fault impact is significant and outputs a sequence of control suggestions for the fault scenario. This provides timely and targeted solutions for fault handling in the distribution network, facilitating the rapid restoration of normal operation. If the scenario difference index does not exceed the preset threshold, it outputs a unified control strategy that includes difference assessment indicators, control suggestions under the fault scenario, and different scenario simulation trajectories for drills and playback. This allows for a comprehensive assessment of the distribution network's operating status and provides data support for subsequent drills and analyses, effectively addressing different operating states of the distribution network.

[0011] As a preferred scheme for the multi-timescale coordinated operation simulation method of distribution networks, the following is provided: The event-based projection of the distribution network operation status, based on the event propagation graph and propagation path set, outputs the evolution sequence of the distribution network operation status over time and the set of activated events in each time step, including: An event-based state evolution mechanism is used to trigger events. After an event is triggered, the state of the corresponding node is updated according to the characteristics of the predecessor events in the propagation path.

[0012] As a preferred scheme for the multi-timescale coordinated operation simulation method of distribution networks, the following is provided: The method of performing event-based deduction of the distribution network operation state based on the event propagation graph and propagation path set, and outputting the evolution sequence of the distribution network operation state over time and the set of activated events in each time step, also includes: After the node status is updated, check whether the node's physical quantities exceed the normal operating threshold. If they do, set the corresponding event to the active state.

[0013] As a preferred scheme for the multi-timescale coordinated operation simulation method of distribution networks, the following is provided: The method of performing event-based deduction of the distribution network operation state based on the event propagation graph and propagation path set, and outputting the evolution sequence of the distribution network operation state over time and the set of activated events in each time step, also includes: After setting the event to the active state, proceed to the next round of event propagation path judgment, until the propagation judgment of all event propagation paths is completed; Throughout the simulation process, discrete time steps are used to recursively calculate within a set maximum simulation time interval until all propagation paths are exhausted or the system state converges; ultimately, the evolution sequence of the distribution network operation state over time and the set of activated events in each time step are generated.

[0014] As a preferred scheme for the multi-timescale coordinated operation simulation method of distribution networks, the following is provided: The process of extending the feedback of the adjusted physical topology map and event propagation path set to multi-scenario comparative simulation, and calculating scenario difference indicators, includes: The feedback process of the adjusted physical topology map and event propagation path set is extended to multi-scenario comparative simulation. The scenarios are divided into normal operation scenarios and fault operation scenarios, and simulation trajectories are generated under normal and fault scenarios. Based on the inference trajectories under normal and fault scenarios, the two inference paths are compared in terms of indicators across different dimensions. Based on the comparison results across multiple dimensions, scenario difference indicators are calculated.

[0015] The beneficial effects of this preferred technical solution are as follows: It extends the feedback process of the adjusted physical topology map and event propagation path set to multi-scenario comparative simulation, dividing the scenarios into normal operation scenarios and fault operation scenarios, enabling a comprehensive comparison of the distribution network's operation under different scenarios. It generates simulation trajectories for normal and fault scenarios, and based on these, compares indicators along the difference dimension, calculating scenario difference indicators, which helps quantify the differences between different scenarios. Through this multi-scenario comparison and indicator calculation, a deeper understanding of the distribution network's operating characteristics under different conditions can be achieved, providing a strong basis for subsequent control decisions and realizing coordinated operation of the distribution network across multiple time scales.

[0016] Secondly, the present invention provides a multi-timescale coordinated operation simulation system for power distribution networks, comprising: The real-time data event propagation module is used to construct event propagation graphs and propagation path sets based on real-time operation data of the power distribution network. The event deduction and status output module is used to perform event-based deduction of the distribution network operation status based on the event propagation graph and propagation path set, and output the evolution sequence of the distribution network operation status over time and the set of activated events in each time step. The structural adjustment determination and generation module is used to call the operation strategy library, determine whether structural adjustment actions need to be performed based on the evolution sequence and the set of activated events, and generate the adjusted physical topology map and event propagation path set. The multi-scenario comparison and simulation and index calculation module is used to extend the feedback process of the adjusted physical topology map and event propagation path set to multi-scenario comparison and simulation, and calculate the scenario difference index. The control strategy output decision module is used to output a control suggestion sequence if the scenario difference index is greater than a preset threshold, and otherwise output a unified control strategy.

[0017] Thirdly, the present invention provides an electronic device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the one or more programs are executed by the one or more processors, the one or more processors implement the multi-timescale coordinated operation simulation method for power distribution networks as described in this invention.

[0018] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the aforementioned method for multi-timescale coordinated operation simulation of a power distribution network.

[0019] The beneficial effects of this invention are as follows: In actual distribution network operation, factors such as the large-scale integration of new energy sources make the operating state complex and changeable, which is difficult to cope with using traditional methods. This invention establishes a two-layer evolution mechanism by uniformly characterizing the causal relationship between abnormal events, state changes, and topology operations, making the operating state a logically coherent dynamic process. This helps to make scheduling decisions more logical and coherent in intelligent distribution scheduling, improving the accuracy and efficiency of scheduling. The introduction of multi-timescale propagation windows and state feedback mechanisms in the modeling can effectively describe fault propagation, structural adjustment, and state response at different time granularities, breaking through the limitations of a single time dimension. This is crucial for emergency operation drills, simulating fault scenarios at different time scales and improving emergency response capabilities. The introduction of a structure-driven state update mechanism realizes a natural closed-loop simulation of topology reconstruction and state evolution. When the actual distribution network topology changes, the system can automatically adjust the state evolution process to ensure stable operation of the distribution network. It can output control suggestion sequences and unified control strategies to provide guidance for fault handling and daily operation. When a fault occurs in the distribution network, it can quickly provide suggestions such as fault location isolation schemes; during normal operation, it can provide optimized unified control strategies. It can be widely used in scenarios such as intelligent power distribution dispatch, emergency operation drills, simulation platform construction and operation and maintenance strategy formulation, providing strong support for the operation analysis of power distribution networks in high-penetration new energy and complex distributed power systems. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is an overall flowchart of the multi-timescale coordinated operation simulation method for power distribution networks provided by the present invention. Detailed Implementation

[0022] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0023] Example 1, referring to Figure 1 This is the first embodiment of the present invention, which provides a method for deducing the coordinated operation of a distribution network across multiple time scales, including: S1: Based on real-time operation data of the power distribution network, construct an event propagation graph and a set of propagation paths; S2: Based on the event propagation graph and propagation path set, perform event-based deduction of the distribution network operation status, and output the evolution sequence of the distribution network operation status over time and the set of activated events in each time step; S3: Call the operation strategy library, determine whether to perform structural adjustment actions based on the evolution sequence and the set of activated events, and generate the adjusted physical topology map and event propagation path set; S4: Extend the feedback process of the adjusted physical topology map and event propagation path set to multi-scenario comparative simulation, and calculate the scenario difference index; S5: If the scenario difference index is greater than the preset threshold, output the regulation suggestion sequence; otherwise, output the unified control strategy.

[0024] It should be noted that through steps S1-S5, the following steps are executed sequentially: constructing an event propagation graph and propagation path set based on real-time operation data of the distribution network; performing event deduction to obtain the evolution sequence and set of activated events; determining whether to execute structural adjustment actions based on the results and generating relevant adjusted content; conducting multi-scenario comparative deduction to calculate scenario difference indicators; and finally outputting corresponding strategies based on the comparison results of scenario difference indicators and preset thresholds. This approach can comprehensively and accurately simulate the operating state of the distribution network under different conditions, promptly identify potential problems, and provide targeted control suggestions or unified control strategies. It helps to achieve coordinated operation of the distribution network across multiple time scales, improve the reliability, stability, and operating efficiency of the distribution network, and ensure the safe and efficient power supply of the power system.

[0025] Example 2, refer to Figure 1 As an embodiment of the present invention, based on the previous embodiment, a method for multi-time-scale coordinated operation simulation of a distribution network is provided, including: In this embodiment, the step S1 above, which involves constructing an event propagation graph and a propagation path set based on real-time operation data of the distribution network, includes: Real-time operation data of the distribution network mainly comes from three system modules: SCADA system, topology management module, and operation alarm and event logging system. The SCADA system provides real-time measurement data, including node voltage, current, load power, etc.; the topology module provides a static network model and real-time switch status; and the operation alarm system records historical and current protection device actions, anomaly logs, and load over-limit information.

[0026] For example, data is typically collected from the substation level to the main station platform via the IEC 104 protocol, IEC 61850, or proprietary interfaces, and can be accessed in real-time or near real-time. Taking a typical 10kV distribution line as an example, its node status data can be collected remotely via feeder terminal units (FTUs) and intelligent ring network cabinets, and updated every 2-5 seconds.

[0027] Based on real-time operation data of the power distribution network, a physical topology diagram, an initial state vector, and an initial set of abnormal events are constructed; based on the physical topology diagram, the initial state vector, and the initial set of abnormal events, an event propagation diagram and a propagation path set are constructed.

[0028] Specifically, constructing a physical topology graph Among them, the node set This includes distribution transformers, high-voltage switches, ring main units, and user load nodes; edge aggregation. Corresponding to conductor segments, cables, and branch connections; attribute set This includes line impedance parameters, switch status, rated capacity, and directional constraints. Whether an edge in the diagram is in an effective on / off state is determined jointly by the switch telemetry data and fault information provided by the SCADA system.

[0029] For example, if the access switch of a branch is in the closed state and the access device is in the "normal" state, then the edge is considered valid; otherwise, the edge will not participate in the subsequent event propagation path analysis; the specific judgment logic is as follows: in, Representing an edge Is it currently accessible? for The remote signaling status of the inter-station switch is uploaded by the substation FTU; A value of 1 indicates a valid connection. A value of 0 indicates that this branch is not currently involved in the spread of the event.

[0030] Constructing the initial system state vector , represented as: in, For nodes The voltage amplitude, in kV, is from SCADA voltage telemetry. For nodes The active power, in kW, is from AMI or load estimates; For nodes The reactive power, measured in kVar; The labels indicate the node's operating status, including whether it is powered on and whether it is affected by protection actions, and are derived from protection records and remote signaling bits.

[0031] Through the Each of them Threshold determination and rate of change identification are performed to form an initial event set. The specific event filtering rules are as follows: like If so, it is marked as a voltage over-limit event; like If so, it is marked as a load mutation event; like And duration > If it does, it is marked as a persistent protection event.

[0032] in, This is within the normal range of voltage amplitude. It is the threshold for changes in active power. This indicates the security duration threshold in the continuous protection event determination.

[0033] Each one that is triggered Each includes its triggering condition, node number, event type, and triggering time.

[0034] It should be noted that the integration of structure recognition and state anomaly recognition creatively generates data through rule inference. and connect it with the structural diagram Simultaneously, the output provides crucial input for constructing structurally constrained event causal propagation paths. In traditional modeling methods, anomaly identification is often a passive response, and event sources and topological constraints are usually modeled decoupled. This embodiment achieves integration from "physical structure—state perception—anomaly identification" through unified modeling and real-time data acquisition.

[0035] Based on the generated distribution network physical topology map Node state vector With the initial set of abnormal events Construct an event propagation graph under structure-state dual constraints. This diagram simulates how operational anomalies such as faults and load fluctuations dynamically propagate along the physical topology of a distribution network. Compared to traditional methods that model propagation based solely on historical events or statistical patterns, this method introduces structural consistency gating functions and state sensitivity constraints for the first time. This ensures that each propagation edge not only exists reasonably within the topology but also realistically reflects the operational dependencies and response characteristics between nodes.

[0036] Specifically, we design a propagation rule generation function that integrates structural topological constraints and state evolution logic. Used for dynamically determining events Can it be triggered? For any two nodes If all of the following constraints are met, the event is considered to be propagable, and an edge is added. to : Structural connectivity constraints: nodes and exist There exist valid connected paths in the equation, and all edges in all paths are connected. (i.e., no circuit breaker tripped); State dependency constraints: Nodes a certain state quantity For nodes status Statistically sensitive dependencies can be evaluated using local linear models; Time window constraint: propagation delay Within the current simulation time window.

[0037] Input the physical topology graph, initial state vector, and initial set of abnormal events into the propagation weight function; output the propagation weight value through the propagation weight function, and include all edges with propagation weight values ​​greater than the threshold into the propagation edge set to form an event propagation graph; construct the propagation path set based on the event propagation graph.

[0038] The expression for the propagation weight function is: in To propagate weight values; , indicating whether there exists a source in the topology from arrive A valid connection; The node state response sensitivity is obtained by fitting the data from the last three months of operation and can be achieved using a linear regression model based on Lasso regularization (to prevent overfitting). for arrive The propagation delay, measured in minutes, is estimated from historical event records; This is a time decay factor (e.g., 0.01~0.1) to control the propagation decay of long-delay paths; The structural conflict index indicates whether there are conditions such as load flow direction conflict, voltage level mismatch, and protection cascading limitations on the current path. This is the conflict penalty coefficient, typically set to 2~10, used to significantly suppress the establishment of unreasonable propagation edges.

[0039] It should be noted that this propagation weighting function introduces a structural conflict regularization term based on topological reachability and state-driven capability. For example, if a path segment has downstream node voltage levels higher than upstream nodes, current directions inconsistent with power flow directions, or crosses protected area boundaries, then... Take 1, thus making The edge is pruned. Compared to existing event modeling methods based on graph neural networks or attention mechanisms, this formula is closer to the physical nature of the power grid, improving interpretability and engineering feasibility.

[0040] All satisfied The edge of the event is incorporated into the spread of the event. ,constitute .in It is the propagation threshold, usually 1. Adjustable.

[0041] Based on the event propagation graph, construct a propagation path set. Each path records the sequence of events within that path. Cumulative delay time and total weight of dissemination .

[0042] It should be noted that this step constructs an event propagation graph that integrates physical constraints and data-driven processes, by introducing a propagation weight function. For the first time, topological connectivity, inter-node state dependencies, time decay, and structural rationality are integrated into a single computational logic, particularly by introducing a structural conflict regularization term. This significantly enhances the engineering controllability and reliability of propagation path construction. This design is particularly suitable for distribution network environments with multiple time scales, strong structural constraints, and coupling of multiple state variables. It not only improves the physical realism of propagation modeling but also provides an interpretable, controllable, and iteratively evolving propagation foundation for subsequent deduction processes. It is the logical starting point and core support for the state-structure linkage deduction mechanism in the entire scheme.

[0043] In another possible implementation, when designing the propagation rule generation function, power flow direction constraints can be considered: the propagation rule can stipulate that events can only propagate along the power flow direction. In a distribution network, power flow typically flows from power source nodes to load nodes. If events propagate in the reverse power flow direction, it may not conform to actual physical laws. For example, when a fault event occurs on the power source side, it usually will not propagate upstream of the power source. In the function design, the power flow direction between nodes can be determined based on the real-time power flow calculation results. Only when the event propagation direction is consistent with the power flow direction is the propagation considered valid.

[0044] In another possible implementation, voltage level constraints can be considered when designing the propagation rule generation function: considering the voltage level differences between nodes, events generally will not propagate from low-voltage level nodes to high-voltage level nodes unless there are special fault conditions (such as misjudgments caused by voltage transformer faults). Node voltage level information can be incorporated into the propagation rule generation function; if there is a propagation path where events propagate from low-voltage level nodes to high-voltage level nodes, then the propagation is deemed invalid.

[0045] In this embodiment, step S2 above, based on the event propagation graph and propagation path set, performs event-based deduction of the distribution network operating state, outputting the evolution sequence of the distribution network operating state over time and the set of activated events in each time step, including: Based on the event propagation graph and propagation path set, the distribution network operation state is extrapolated based on events in conjunction with the initial state vector, and the evolution sequence of the distribution network operation state over time and the set of activated events in each time step are output. Specifically, an event-based state evolution mechanism is used to trigger events; After an event is triggered, the state of its corresponding node is updated according to the characteristics of the preceding events in the propagation path; After a node is updated, if any physical quantity of the node exceeds the normal operating threshold, the corresponding event will be set to active. Once set to active state, proceed to the next round of event propagation path judgment until all event propagation path judgments are completed.

[0046] Furthermore, the state update formula for a node is: in For nodes At any moment The state increment; For the event All precursor events of propagation; For the propagation path arrive The intensity of transmission; As a delay-sensitive factor, controlling the weakening effect of propagation delay; for arrive The delay in the spread; For the maximum simulation time window; It is a structural interference factor; This is the structural disturbance regularization coefficient, used to suppress unreasonable paths; For nodes The state corresponding to the propagation delay; For nodes The state before the update.

[0047] Delay adjustment item This is used to model the intensity attenuation of events with different propagation delays along a propagation path. In reality, short-delay events (such as protection trips) are more sudden and impactful than long-delay events (such as voltage fluctuations), therefore, it is necessary to model the intensity attenuation of events. Time-weighted summation reflects physical laws; express spread to The degree of structural disturbance along the propagation path, for example, if there are behaviors such as frequent switch switching, load node connection / disconnection, intermittent connection of distributed power sources, and reverse power flow in the propagation path, the system assigns... The value represents the degree of path instability, suppressing the unreasonable amplification of states by such paths. This regularization term can be obtained by analyzing the operational behavior logs from the past two weeks, calculating the frequency of structural fluctuations on the path, and then normalizing it. For example, if there were more than 3 switch opening and closing operations on this path in the past hour, Can be set to If the path remains stable and unchanged, then .

[0048] The state update is an incremental process. The state sequence update for each node within the simulation period is as follows: After the update, a status limit check is required. If the node... Any physical quantity (such as) , , If the normal operating threshold is exceeded (which can be set according to technical specifications, such as voltage deviation ±7%, load overload 110%, etc.), then the corresponding event will be triggered. Set to active state and proceed to the next round of propagation path determination.

[0049] The entire simulation is conducted within a time frame. The process is repeated in discrete time steps (e.g., 1 minute) until all propagation paths are exhausted or the system state converges, during which a state trajectory is formed. .

[0050] Finally, an evolution sequence of the distribution network's operating status over time is generated. Each time step contains the state of all nodes in the network; the set of events activated within each time step. .

[0051] It should be noted that this step deeply integrates the event graph-driven mechanism with the physical evolution process of the state, proposing an interpretable and evolvable state update method. By introducing delay adjustment terms and structural disturbance regularization terms, it not only characterizes the multi-timescale impact of abnormal events on the operating state in actual distribution networks, but also effectively avoids the risk of misjudgment caused by unreasonable propagation paths in conventional modeling. Especially in modern distribution systems with a large number of switchable topologies and frequent changes in distributed power sources, this model achieves system modeling of "path instability" through path disturbance identification and propagation weight correction. It is a key design for achieving highly reliable and physically consistent deduction, and also the dynamic driving core of the state-structure coupled evolution mechanism in the entire scheme.

[0052] In another possible implementation, when performing event-based extrapolation of the distribution network's operating status, the distribution network can be divided into different scales, such as high-voltage, medium-voltage, and low-voltage distribution networks, and corresponding event propagation models can be constructed for each. Models at different scales can interact through boundary conditions and data interfaces to achieve multi-scale collaborative extrapolation. For example, fault events in the high-voltage distribution network can affect the operating status of the medium-voltage distribution network through boundary nodes, and load changes in the medium-voltage distribution network can also be fed back into the high-voltage distribution network model.

[0053] In another possible implementation, when performing event-based simulations of the distribution network's operating status, the scheduling scheme can be adjusted and optimized in real time based on the simulation results. For example, in fault scenarios, the impact of different load reduction schemes and distributed generation access schemes on the distribution network's operating status can be analyzed, and the optimal scheduling strategy can be selected to restore the distribution network's normal operation at the lowest cost. Simultaneously, under normal operating conditions, the output allocation of distributed generation sources and the balanced distribution of loads can also be dynamically adjusted based on the simulation results.

[0054] In this embodiment, step S3 above calls the operation strategy library to determine whether a structural adjustment action needs to be performed based on the evolution sequence and the set of activated events, generating the adjusted physical topology map and event propagation path set, including: The operation strategy library stores executable structural adjustment operations in the form of rule tables. Each rule table contains applicable conditions, topology impact amount, and maximum allowed number of responses. The operation strategy library searches for all structured response rules that meet the specified form based on the operation type of each activation event; The update operation is performed according to the structural response rules. After the update operation is performed, the propagation graph is corrected by the local update function of the propagation path, and the physical topology graph and propagation path set with structural consistency are reconstructed.

[0055] For example, if pu and neighboring regions exist If pu is selected, a "load transfer" operation is performed; among which, This represents the voltage amplitude at node j at time step t. This represents the voltage amplitude at node k at time step t.

[0056] like And continue If min, then perform the "branch isolation" operation; This indicates the running status label of node j at time step t.

[0057] The result of executing each rule is Modify the properties of a set of edges For example, a switch changes from closed to open, or a branch is reconfigured and connected. Structural changes are represented by the difference in the edge state matrix. This means that the system only updates edges that have changed.

[0058] Assume that the voltage at node j at the end of the feeder is abnormal ( pu), event Once activated, the system checks the rule base and finds a matching rule: like And neighboring nodes satisfy It allows "loop reconfiguration" to switch to the neighboring cell feeder.

[0059] The system performs this operation to update the edges. and in Adding branch roads .

[0060] Furthermore, the expression for the local update function of the propagation path is: in This is the revised set of event propagation paths; This represents the set of propagation paths leading to failure due to structural discontinuity. This refers to the newly added propagation path due to the new connection of the structure.

[0061] It should be noted that the local update function of the propagation path avoids the high complexity of full graph reconstruction. It only replaces the parts of the propagation path that have changed, and retains or slightly adjusts the propagation strength, time delay and other attributes, thus maintaining the continuity of the system's evolution trajectory.

[0062] Finally, the physical topology diagram after structural adjustment is obtained. Event propagation path set after structural consistency correction .

[0063] In this embodiment, the feedback process of the adjusted physical topology map and event propagation path set in step S4 above is extended to multi-scenario comparison and deduction, and the scenario difference index is calculated, including: The feedback process of the adjusted physical topology map and event propagation path set is extended to multi-scenario comparative simulation, and divided into normal operation scenarios and fault operation scenarios; the simulation trajectory under the normal scenario is obtained. ( ); Deduction trajectory under fault scenarios (Using the identification in S1) );in, This represents the set of system state vectors under the normal simulation scenario at time step t. This represents the physical topology diagram under normal scenario at time step t. This represents the set of event propagation paths in a normal simulation scenario at time step t. This represents the set of system state vectors under the control simulation scenario at time step t. This indicates the physical topology diagram of the simulation scenario at time step t. This represents the set of event propagation paths in the control simulation scenario at time step t.

[0064] These two trajectories can be generated in parallel, using the same initial structure diagram and initial state vector. The only difference is whether there is an initial disturbance event.

[0065] After performing simulations of two scenarios in parallel, the metrics of the two simulation paths are compared, with a particular focus on the following three dimensions of difference: state evolution deviation. Structural response differences Control strategy differences: Structure operation set and The coverage and operation timing differences.

[0066] Define scenario difference indicators , represented as: in, Represents the Euclidean difference between the states in two scenarios; This indicates the difference in the number of structural changes; The set similarity between sequences of structural operations. As the event's impact attenuation factor, This is the topological conflict penalty coefficient. Update the smoothing factor for the state. , , Configured according to operating strategy preferences.

[0067] In another possible implementation, when calculating the scenario difference index, the operating frequency of control devices (such as switches, capacitors, etc.) under normal and fault scenarios can also be statistically analyzed. An increase in the control operating frequency may indicate system instability or frequent adjustments to the control strategy. The difference or ratio of the control device operating frequencies under the two scenarios can be calculated to measure the difference in operating frequency of the control strategy.

[0068] In another possible implementation, when calculating the scenario difference index, critical lines in the distribution network can be identified, and their importance differences under normal and fault scenarios can be calculated. A power flow transfer-based method can be used to calculate the power flow transfer coefficient of each line under different scenarios. The larger the power flow transfer coefficient, the higher the importance of the line in maintaining stable system operation. By comparing the power flow transfer coefficients of critical lines in the two scenarios, the change in their importance can be assessed.

[0069] In this embodiment, if the scenario difference index in step S5 is greater than a preset threshold, a regulation suggestion sequence is output; otherwise, a unified control strategy is output, including: When the scenario difference index Exceeding the preset threshold The system will mark faults as having a significant impact and output a sequence of control recommendations for the fault scenario. This includes: fault location isolation schemes; local reconstruction strategies; suggestions for adjusting the control window of the fault-induced area (such as avoiding access fluctuation sources); and manual intervention window when the system fails to self-heal.

[0070] Otherwise, the output includes difference assessment indicators, control suggestions under fault scenarios, and a unified control strategy for different scenario simulation trajectories used in drills and replays, indicating that the current network structure has a certain degree of robustness and does not require differentiated responses.

[0071] In another possible implementation, the preset thresholds can be dynamically adjusted based on the real-time operating status of the distribution network. For example, during peak load periods, due to the greater operating pressure on the system, the thresholds for some differential indicators can be appropriately lowered to more rigorously monitor the system's operation. Conversely, during off-peak load periods, the thresholds can be appropriately raised to reduce unnecessary warnings and operations.

[0072] In another possible implementation, the preset threshold can be set based on the distribution network operator's risk appetite and tolerance for different operating conditions. If the operator prioritizes system reliability and security and is willing to accept lower risks, the threshold can be set more strictly, meaning that even a small difference in the indicator value will trigger corresponding early warning or control measures. Conversely, if the operator can accept system fluctuations and risks to a certain extent in order to pursue higher economic benefits, the threshold can be appropriately relaxed. For example, for voltage quality difference indicators, an operator with a lower risk appetite might set the threshold at ±3% voltage deviation, while an operator with a higher risk appetite might relax the threshold to ±5%.

[0073] Example 3: The above is an illustrative scheme of the distribution network multi-timescale coordinated operation simulation method of this embodiment. It should be noted that the technical solution of the distribution network multi-timescale coordinated operation simulation system and the technical solution of the above-described distribution network multi-timescale coordinated operation simulation method belong to the same concept. Details not described in detail in the technical solution of the distribution network multi-timescale coordinated operation simulation system in this embodiment can be found in the description of the technical solution of the above-described distribution network multi-timescale coordinated operation simulation method.

[0074] This embodiment also provides a multi-timescale coordinated operation simulation system for distribution networks, including: The real-time data event propagation module is used to construct event propagation graphs and propagation path sets based on real-time operation data of the power distribution network. The event deduction and status output module is used to perform event-based deduction of the distribution network operation status based on the event propagation graph and propagation path set, and output the evolution sequence of the distribution network operation status over time and the set of activated events in each time step. The structural adjustment determination and generation module is used to call the operation strategy library, determine whether structural adjustment actions need to be performed based on the evolution sequence and the set of activated events, and generate the adjusted physical topology map and event propagation path set. The multi-scenario comparison and simulation and index calculation module is used to extend the feedback process of the adjusted physical topology map and event propagation path set to multi-scenario comparison and simulation, and calculate the scenario difference index. The control strategy output decision module is used to output a control suggestion sequence if the scenario difference index is greater than a preset threshold, and otherwise output a unified control strategy.

[0075] This embodiment also provides an electronic device applicable to the multi-timescale coordinated operation simulation method of distribution networks, including: The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement the multi-timescale coordinated operation simulation method for power distribution networks proposed in the above embodiments.

[0076] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements the multi-timescale coordinated operation simulation method for distribution networks proposed in the above embodiments.

[0077] The storage medium proposed in this embodiment and the multi-timescale coordinated operation simulation method for power distribution networks proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0078] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for predicting the coordinated operation of a distribution network across multiple time scales, characterized in that, include: Based on real-time operation data of the power distribution network, an event propagation graph and a set of propagation paths are constructed. Based on the event propagation graph and propagation path set, the operation status of the distribution network is extrapolated based on events, and the evolution sequence of the operation status of the distribution network over time and the set of activated events in each time step are output. The operation strategy library is invoked to determine whether structural adjustment actions need to be performed based on the evolution sequence and the set of activated events, and to generate the adjusted physical topology and event propagation path set. The feedback process of the adjusted physical topology map and event propagation path set is extended to multi-scenario comparative simulation, and scenario difference indicators are calculated; If the scenario difference index is greater than the preset threshold, a regulation suggestion sequence will be output; otherwise, a unified control strategy will be output.

2. The method for multi-timescale coordinated operation simulation of a distribution network as described in claim 1, characterized in that, The construction of the event propagation graph and propagation path set based on real-time operation data of the distribution network includes: Based on real-time operation data of the power distribution network, a physical topology diagram, an initial state vector, and an initial set of abnormal events are constructed. Based on the physical topology graph, initial state vector, and initial abnormal event set, a propagation rule generation function is designed to determine whether an event can propagate between nodes. The physical topology graph, initial state vector, and initial abnormal event set are input into the propagation weight function, which outputs the propagation weight value. Edges with propagation weight values ​​greater than a threshold are included in the propagation edge set to form the event propagation graph. Based on the event propagation graph, a propagation path set is constructed, with each path recording the event sequence, cumulative delay time, and total propagation weight.

3. The method for multi-timescale coordinated operation simulation of a distribution network as described in claim 2, characterized in that, If the scenario difference index is greater than a preset threshold, a regulation suggestion sequence is output; otherwise, a unified control strategy is output, including: The calculated scenario difference index is compared with a preset threshold. If the scenario difference index exceeds the preset threshold, the system determines that the fault has a significant impact and outputs a sequence of control suggestions for the fault scenario. If the scenario difference index does not exceed the preset threshold, the output includes the difference assessment index, control suggestions under the fault scenario, and unified control strategy for different scenario simulation trajectories used for drills and playback.

4. The method for multi-timescale coordinated operation simulation of a distribution network as described in claim 3, characterized in that, The event-based projection of the distribution network operation status, based on the event propagation graph and propagation path set, outputs the evolution sequence of the distribution network operation status over time and the set of activated events in each time step, including: An event-based state evolution mechanism is used to trigger events. After an event is triggered, the state of the corresponding node is updated according to the characteristics of the predecessor events in the propagation path.

5. The method for multi-timescale coordinated operation simulation of a distribution network as described in claim 4, characterized in that, The method of performing event-based deduction of the distribution network operation state based on the event propagation graph and propagation path set, and outputting the evolution sequence of the distribution network operation state over time and the set of activated events in each time step, also includes: After the node status is updated, check whether the node's physical quantities exceed the normal operating threshold. If they do, set the corresponding event to the active state.

6. The method for multi-timescale coordinated operation simulation of a distribution network as described in claim 5, characterized in that, The method of performing event-based deduction of the distribution network operation state based on the event propagation graph and propagation path set, and outputting the evolution sequence of the distribution network operation state over time and the set of activated events in each time step, also includes: After setting the event to the active state, proceed to the next round of event propagation path judgment, until the propagation judgment of all event propagation paths is completed; Throughout the simulation process, discrete time steps are used to recursively calculate within a set maximum simulation time interval until all propagation paths are exhausted or the system state converges; ultimately, the evolution sequence of the distribution network operation state over time and the set of activated events in each time step are generated.

7. The method for multi-timescale coordinated operation simulation of a distribution network as described in claim 6, characterized in that, The process of extending the feedback of the adjusted physical topology map and event propagation path set to multi-scenario comparative simulation, and calculating scenario difference indicators, includes: The feedback process of the adjusted physical topology map and event propagation path set is extended to multi-scenario comparative simulation. The scenarios are divided into normal operation scenarios and fault operation scenarios, and simulation trajectories are generated under normal and fault scenarios. Based on the inference trajectories under normal and fault scenarios, the two inference paths are compared in terms of indicators across different dimensions. Based on the comparison results across multiple dimensions, scenario difference indicators are calculated.

8. A multi-timescale coordinated operation simulation system for a distribution network, using the method described in any one of claims 1 to 7, characterized in that, include: The real-time data event propagation module is used to construct event propagation graphs and propagation path sets based on real-time operation data of the power distribution network. The event deduction and status output module is used to perform event-based deduction of the distribution network operation status based on the event propagation graph and propagation path set, and output the evolution sequence of the distribution network operation status over time and the set of activated events in each time step. The structural adjustment determination and generation module is used to call the operation strategy library, determine whether structural adjustment actions need to be performed based on the evolution sequence and the set of activated events, and generate the adjusted physical topology map and event propagation path set. The multi-scenario comparison and simulation and index calculation module is used to extend the feedback process of the adjusted physical topology map and event propagation path set to multi-scenario comparison and simulation, and calculate the scenario difference index. The control strategy output decision module is used to output a control suggestion sequence if the scenario difference index is greater than a preset threshold, and otherwise output a unified control strategy.

9. An electronic device, characterized in that, include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores computer-executable instructions that, when executed by a processor, implement the steps of the method according to any one of claims 1 to 7.

Images