An event-driven master-slave coordination incremental power flow simulation method

Abstract

This invention discloses an event-driven primary-distribution cooperative incremental power flow simulation method, specifically relating to the field of power system dynamic modeling and simulation technology. The method receives discrete driving events, generates event tuples, merges them to form event chains, establishes a primary-distribution joint influence diagram, and determines the initial influence domain. Based on the topology increment and injection increment within the influence domain, it obtains the predicted boundary exchange power change, determines the boundary penetration risk level by combining the minimum electrical adjacency level and the boundary coupling path propagation depth, and divides the event chain into local event chains or cross-domain event chains. It only performs topology increment updates, injection increment updates, and local incremental power flow simulation within the influence domain, and performs primary-distribution boundary cooperative correction when the boundary active power mismatch, boundary reactive power mismatch, or boundary voltage deviation exceeds a threshold. If necessary, it expands the influence domain according to the electrical adjacency level and repeats the solution to achieve fast and accurate simulation of primary-distribution cooperative power flow under discrete event scenarios.

Description

Technical Field

[0001] This invention relates to the field of dynamic modeling and simulation technology of power systems, and more specifically, to an event-driven primary-distributor coordinated incremental power flow simulation method. Background Technology

[0002] With the large-scale integration of distributed photovoltaic systems, energy storage devices, flexible loads, tie switches, and reactive power compensation equipment on the distribution side, the operating status of the distribution network changes frequently, and the boundary coupling relationship between the distribution network and the transmission network is significantly enhanced. Existing dispatching master stations or monitoring and control and data acquisition systems typically continuously receive multi-source heterogeneous data such as switch measurements, boundary measurements, distributed power source monitoring, and load acquisition, and need to conduct equipment status analysis, operation mode adjustment, and power flow simulation based on this data. Especially in scenarios such as feeder transfer, tie switch switching, sectional switch opening and closing, distributed power source output fluctuations, and reactive power compensation device switching, discrete events often exhibit characteristics of continuous arrival, interrelation, and cross-boundary propagation, which places higher demands on the real-time performance and accuracy of simulations in operational status analysis.

[0003] Most existing power flow analysis methods rely on static modeling of the entire network and repeated solutions across the entire network. When the switching state or node injection changes, it is usually necessary to rebuild or update a large-scale network model and perform a complete power flow calculation again. Although this method can obtain calculation results, it leads to significant overhead from repeated modeling and solving in high-frequency event-driven scenarios, making it difficult to meet the requirements of online analysis and rapid simulation. Especially when multiple related events arrive consecutively within a short time window, solving each event independently can easily lead to repeated calculations of the same disturbance paths, increasing the number of boundary coordination attempts, expanding the scope of the solution object, and decreasing computational efficiency.

[0004] On the other hand, traditional methods of separating the main grid and distribution network modeling typically employ equivalent or fixed boundary conditions to handle the coupling relationship between the main grid and distribution network. While this approach has some applicability in steady-state and low-frequency variation scenarios, it often fails to reflect the dynamic impact of topology reconfiguration, injected mutations, and disturbances propagating along the main-distribution coupling path near boundary nodes on boundary exchange power, boundary voltage, and the operating status of adjacent areas. When the distribution network model is updated only locally without considering boundary state verification and collaborative correction, inconsistencies can easily arise between the local solution results and actual boundary measurements or the boundary states of the higher-level dispatching system, thus affecting the reliability of the simulation results.

[0005] Furthermore, while some existing local solution methods attempt to narrow the solution scope, they typically lack a unified organizational mechanism for the correlation of discrete events. This prevents them from chaining multiple events that are temporally close, topologically related, and propagating along the same boundary direction. They also lack an effective mechanism to determine whether the influence of an event will penetrate the primary and secondary boundaries, making it difficult to distinguish between local events requiring only local solution and cross-domain events requiring cross-domain linkage. For local solution results that do not meet accuracy requirements, existing methods generally lack a closed-loop processing mechanism to gradually expand the influence domain according to electrical adjacency layers, resulting in either an excessively large solution scope with high computational costs, or an excessively small solution scope with insufficient boundary consistency.

[0006] Therefore, there is an urgent need in this field for a unified dynamic modeling, local incremental power flow simulation, on-demand boundary collaborative correction, and layer-by-layer expansion solution method for main-distribution systems in high-frequency discrete event-driven scenarios. Summary of the Invention

[0007] To overcome the aforementioned deficiencies in existing technologies, this invention provides an event-driven primary-distributor collaborative incremental power flow simulation method. By constructing event chains and a joint primary-distributor influence diagram, it performs unified dynamic modeling of discrete operating events and obtains the predicted boundary exchange power changes based on topology increments and injection increments within the influence domain, determining the boundary penetration risk level and achieving separate processing of local and cross-domain event chains. Furthermore, it performs topology increment updates, injection increment updates, and local incremental power flow simulation only on the influence domain, and implements primary-distributor boundary collaborative correction and progressive solution by expanding the influence domain according to the electrical adjacency layer when necessary, thereby addressing the problems mentioned in the background technology.

[0008] To achieve the above objectives, the present invention provides the following technical solution: An event-driven primary-distribution collaborative incremental power flow simulation method includes the following steps: acquiring a transmission network model, a distribution network model, a primary-distribution boundary node set, and node voltages, branch power flows, and boundary exchange states before an event occurs; receiving discrete events and generating event tuples, merging them into event chains according to a preset time window and equipment topology reachability; establishing a primary-distribution joint influence diagram based on the event chains, and determining the initial influence domain based on the directly affected nodes of the event chains, branches affected by topology changes, and adjacent nodes and branches connected along the power flow propagation direction; and determining the initial influence domain based on the minimum electrical adjacency level from the initial influence domain to the primary-distribution boundary node set, the boundary coupling path propagation depth, and the topology increment and injection increment within the influence domain. The predicted boundary exchange power change is used to determine the boundary penetration risk level, and the event chain is divided into local event chains or cross-domain event chains. Topology incremental update and injection incremental update are performed only within the influence domain, while maintaining the pre-event state of non-influence domains for local incremental power flow simulation. When any of the boundary active power mismatch, boundary reactive power mismatch, and boundary voltage deviation does not meet the corresponding threshold, primary and secondary boundary collaborative correction is performed. If the stopping condition is still not met after correction, the influence domain is expanded outward according to the electrical adjacency layer, and the boundary penetration risk level determination, incremental update, local incremental power flow simulation, and primary and secondary boundary collaborative correction are repeated until the stopping condition is met, and then the primary and secondary collaborative power flow simulation results after the event are output.

[0009] As a further aspect of the present invention, the event tuple includes at least event type, occurrence location, acting device, occurrence time, change value, and data confidence level. The event type includes topology events and injection events. The topology events include tie switch closing events, sectionalizing switch opening events, sectionalizing switch closing events, and feeder transfer events. The injection events include distributed power source active power output change events, distributed power source reactive power output change events, load change events, and reactive power compensation device switching events. Boundary measurement values ​​are only used for verification of local incremental power flow simulation results and are not used as constituent elements of the event chain.

[0010] As a further aspect of the present invention, the step of grouping event elements into event chains according to a preset time window and the topological reachability of the equipment includes: grouping event elements whose occurrence time difference does not exceed the preset time window and which act on the same feeder, the same branch switching channel, or adjacent electrical equipment into the same candidate event chain, and then determining the candidate event chain as the target event chain based on the causal continuity, power flow transmission continuity, and common action boundary direction between the event elements.

[0011] As a further aspect of the present invention, a main-distribution joint influence diagram is established before constructing the initial influence domain. The nodes of the main-distribution joint influence diagram include main grid nodes, distribution network nodes, boundary nodes, switch action nodes, and distributed power generation nodes or load nodes. The edges of the main-distribution joint influence diagram include transmission line edges, distribution branch edges, main-distribution boundary coupling edges, and event causal edges. The initial influence domain is determined based on the directly acting nodes of the event chain, the branches affected by topology changes, and the adjacent nodes and branches connected to the directly acting nodes along the power flow transmission direction.

[0012] As a further aspect of the present invention, the boundary penetration risk level is determined by a combination of the following factors: the minimum electrical adjacency level from the event chain's action area to the set of primary and secondary boundary nodes, the propagation depth of the event chain along the boundary coupling path, the predicted change in boundary exchange power caused by topological increments within the influence domain, and the predicted change in boundary exchange power caused by injection increments within the influence domain; when the boundary penetration risk level is lower than a preset boundary threshold, the event chain is determined to be a local event chain; otherwise, it is determined to be a cross-domain event chain.

[0013] As a further aspect of the present invention, the predicted boundary exchange power change is obtained by updating the local branch current relationship based on the change of switch state within the influence domain, and combining the active power injection increment and reactive power injection increment of the changed node with the directional branch current relationship corresponding to the boundary support branch to obtain the predicted results of the active power exchange change and reactive power exchange change at the boundary node.

[0014] As a further aspect of the present invention, the primary-distribution boundary collaborative correction includes: generating boundary equivalent injection correction and boundary voltage correction based on the boundary active power mismatch, boundary reactive power mismatch, and boundary voltage deviation between the local incremental power flow simulation results and the boundary measurement values ​​or the upper-level scheduling boundary state; and updating the equivalent injection parameters and boundary node voltages of the boundary nodes within the influence domain based on the boundary equivalent injection correction and boundary voltage correction, and then re-executing the local incremental power flow simulation.

[0015] As a further aspect of the present invention, the incremental injection update includes: updating the active power injection parameters and reactive power injection parameters of the source nodes and load nodes that have undergone injection changes within the influence domain; keeping the injection parameters of the nodes that have not undergone injection changes within the influence domain unchanged before the event; and keeping the node voltage, branch power flow and boundary exchange state of the nodes outside the influence domain unchanged before the event.

[0016] As a further aspect of the present invention, the stopping conditions are that the boundary active power mismatch is not greater than a first preset threshold, the boundary reactive power mismatch is not greater than a second preset threshold, and the boundary voltage deviation is not greater than a third preset threshold. If any condition is not met, it is determined that the stopping condition has not been met.

[0017] As a further aspect of the present invention, the step of expanding the influence domain outward according to the electrical adjacency layer includes: incorporating the next layer nodes and branches that are directly adjacent to the boundary support branch, the maximum mismatch propagation path, or the node with the maximum boundary voltage deviation outside the current influence domain into the influence domain; updating the local branch current relationship, the boundary coupling path set, and the variable node injection parameters again within the expanded influence domain; and repeatedly performing local incremental power flow simulation and main distribution boundary collaborative correction.

[0018] As a further aspect of this invention, the event-driven primary-distribution collaborative incremental power flow simulation method can be deployed in the power operation analysis and collaborative scheduling environment of an industrial internet platform. The industrial internet platform performs unified access and unified calculation scheduling of the transmission network model, distribution network model, primary-distribution boundary node set, and local incremental power flow simulation results. Combined with the industrial internet security management mechanism, hierarchical permission control is implemented for the event tuple generation and influence domain expansion solution process. When abnormal confidence of discrete event data or continuous exceedance of boundary mismatch caused by cross-domain event chain is detected, the industrial internet platform can trigger the corresponding simulation task's security review, incremental recalculation, or result isolation, thereby preventing abnormal event data, erroneous boundary states, or unreliable simulation results from continuing to propagate in the primary-distribution collaborative analysis link, further improving the efficiency, boundary consistency, and operational security of the primary-distribution collaborative incremental power flow simulation of this invention in the industrial internet platform scenario.

[0019] The technical effects and advantages of the event-driven master-distributor cooperative incremental power flow simulation method of the present invention are as follows: This invention organizes discrete events into event tuples and constructs event chains, and determines the initial influence domain by combining the principal-partition joint influence diagram. This enables event-driven dynamic modeling in principal-partition coupled operation scenarios, and can uniformly handle related events that are close in time, topologically related, and propagate along the same boundary direction.

[0020] This invention determines local and cross-domain event chains based on the risk level of boundary penetration, and performs topology incremental updates and injection incremental updates only within the affected domain. This maintains the pre-inheritance state of events in non-affected domains, reduces the participation of irrelevant nodes and branches in the calculation, thereby reducing the computational scale and improving the efficiency of online simulation.

[0021] This invention uses a boundary verification and primary-feeder boundary collaborative correction mechanism to generate equivalent injection correction and boundary voltage correction based on boundary active power mismatch, boundary reactive power mismatch, and boundary voltage deviation. This can improve the consistency between local incremental power flow simulation results and actual boundary states, and enhance the accuracy and reliability of simulation results in primary-feeder coupling scenarios.

[0022] This invention expands the influence domain outward from the electrical adjacent layer when the stopping condition is not met, and repeatedly performs risk level determination, incremental update, local incremental power flow simulation and primary-secondary boundary collaborative correction, forming a closed-loop processing mechanism that expands layer by layer and converges gradually. It can ensure boundary accuracy requirements while taking into account computational efficiency, and has good engineering application value. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the overall process of an event-driven master-distributor cooperative incremental power flow simulation method according to the present invention.

[0024] Figure 2 This is a schematic diagram of the combined influence of the main components in Embodiment 1 of the present invention. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] Example 1 This embodiment illustrates an event-driven incremental power flow simulation method based on main-distribution coordination under the condition of high load and rapid fluctuations in distributed photovoltaic output during summer. This embodiment is executed in the computing environment of a distribution automation master station or a monitoring, control, and data acquisition system. The master station preloads the transmission network model, distribution network model, set of main-distribution boundary nodes, and current node voltages, branch power flows, and boundary exchange states, and continuously receives multi-source heterogeneous discrete operating data streams from switchgear, distributed power source monitoring units, boundary measurement units, and load acquisition terminals. The master station performs time-scale unification, anomaly removal, data confidence assessment, and preprocessing on the data streams to generate the input data required for subsequent event chain construction, main-distribution joint influence diagram establishment, initial influence domain identification, boundary penetration risk level determination, and local incremental power flow simulation.

[0027] This embodiment constructs an event chain and establishes a primary-secondary joint influence diagram to perform event-driven dynamic modeling of the primary-secondary coupled operating state, and executes incremental power flow simulation within the influence domain, such as... Figure 1 As shown, the method flow of this embodiment includes, in sequence, obtaining the running status, receiving discrete events, generating event tuples, constructing event chains, establishing a primary-secondary joint influence diagram, determining the initial influence domain, determining the boundary penetration risk level, updating the topology incrementally, injecting incremental updates, simulating local incremental power flow, verifying the boundary, coordinating the primary-secondary boundary correction, expanding the electrical adjacency layer, and outputting the results.

[0028] This embodiment selects four 10kV feeders of a 110kV substation in a certain region as the execution scenario. The substation is connected to the upstream transmission network through main-distribution boundary nodes B1 and B2, and the 10kV side is powered by two 50MVA main transformers. The four feeders are designated as F1, F2, F3, and F4. A tie switch L12 is installed between F1 and F2, a sectionalizing switch S27 is installed in the middle of F2, and the end of F2 is connected to a rooftop distributed photovoltaic power station PV2 with a rated capacity of 3.2MW. A set of switchable reactive power compensation device C23 is installed upstream of F2. The distribution network includes 58 nodes, 57 branches, 12 sectionalizing switches, 3 distributed power sources, and 1 set of reactive power compensation device. At the current moment, the total active load of the distribution network is 18.6MW, the total reactive load is 6.8Mvar, the boundary active power exchange is 16.4MW, the boundary reactive power exchange is 5.9Mvar, the boundary node voltage is 1.012pu, the actual active power output of PV2 is 2.7MW, the reactive power output is 0.12Mvar, the voltage of the most unfavorable node at the end of F2 is 0.964pu, and the utilization rate of the heaviest-loaded branch current is 82.3%. The above quantities, as the current operating state, constitute the initial basis for the incremental power flow simulation of this embodiment. The master station uses the node voltage, branch power flow, and boundary exchange state before the event as the reference benchmark for the subsequent non-affected domain inherited state.

[0029] At 14:12:03, the master station received the state change information of the tie switch L12, forming the first event tuple E1. The event type of E1 is a topology event, the location of the event is the tie point between F1 and F2, the device involved is L12, the time of the event is 14:12:03.214, the change value is the switch state changing from open to closed, and the data confidence level is 0.996.

[0030] At 14:12:04, the master station received the state change information of the segment switch S27, forming the second event tuple E2. The event type of E2 is a topology event, the occurrence location is the middle of F2, the acting device is S27, the occurrence time is 14:12:04.021, the change value is the switch state changing from closed to open, and the data confidence level is 0.998.

[0031] At 14:12:07, the master station received the power output change information from PV2, forming the third event tuple E3. The event type of E3 is an injection event, the location of the event is the distributed power access point at the end of F2, the affected device is PV2, the time of the event is 14:12:07.506, the change value is that the active power output decreased from 2.7MW to 0.9MW and the reactive power output decreased from 0.12Mvar to 0.05Mvar, and the data confidence level is 0.991.

[0032] At 14:12:08, the master station received the boundary measurement values ​​of boundary node B1 and the boundary status data of the upper-level scheduling node. This data is only used as the boundary verification data M1 for subsequent local incremental power flow simulation results and as the basis for primary-distributor boundary collaborative correction. It does not generate event tuples or participate in event chain construction.

[0033] The master station constructs event chains for discrete events. This construction includes two stages: candidate event chain formation and target event chain determination. In the first stage, the master station uses a preset time window of 10 seconds and, based on equipment topology reachability, groups event elements whose occurrence time difference does not exceed the preset time window and which act on the same feeder, the same branch switching channel, or adjacent electrical equipment into the same candidate event chain. In the second stage, the master station further determines the candidate event chain as the target event chain based on the causal continuity, power flow continuity, and common action boundary direction between the event elements. According to the above rules, E1, E2, and E3 all occur within the same preset time window and jointly act on the feeder reconfiguration path between F1 and F2 and the subsequent power flow propagation path. E1 and E2 constitute a causal continuity operation for interconnection reconfiguration, and the distributed power output change corresponding to E3 continues to propagate along the reconfigured power supply path to the boundary node set. Therefore, the master station first groups E1, E2, and E3 into the same candidate event chain, and then, combining their common action boundary direction pointing to the boundary node sets B1 and B2, determines it as the target event chain C1.

[0034] Before constructing the initial influence domain, the main site establishes a joint influence map of the main and supporting systems, such as... Figure 2 As shown, the main and distribution network joint influence diagram includes main network nodes, distribution network nodes, boundary nodes, switch action nodes, and distributed power generation nodes or load nodes, and establishes correlation relationships through transmission line edges, distribution branch edges, main and distribution boundary coupling edges, and event causal edges.

[0035] On the joint influence map, the main station first determines the initial influence domain D0; when the stopping condition is still not met after boundary collaborative correction, the extended influence domain D1 is formed by expanding outward according to the electrical adjacency layer.

[0036] Specifically, an event causal edge is established between E1 and tie switch L12, between E2 and sectionalizing switch S27, and between E3 and PV2 node; boundary nodes B1 and B2 are connected to the distribution network through the main-distribution boundary coupling edge. The master station determines the initial influence domain D0 based on the directly affected nodes of event chain C1, the branches affected by topology changes, and the adjacent nodes and branches connected to the directly affected nodes along the power flow direction. D0 contains 18 nodes and 17 branches, covering the tie-forward path between F1 and F2, the photovoltaic access branch at the end of F2, and the boundary support branch coupled to it, but does not include the irrelevant areas of F3 and F4.

[0037] After obtaining the initial influence domain D0, the master station first performs topology incremental update and injection incremental update within the influence domain. For the topology change part, after the tie switch L12 changes from open to closed, a new tie path is formed between F1 and F2; after the sectionalizing switch S27 changes from closed to open, the original power supply branch in the middle of F2 is cut off. The master station establishes a local branch current relationship matrix B within D0. loc,0 The matrix is ​​structured with branches within the influence domain as rows and nodes within the influence domain (excluding local reference nodes) as columns. In this embodiment, B... loc,0 The dimensions are 17×17. The main station constructs a temporary vector by subtracting the associated column vectors of the two ends of the closed branch L12 in the local branch current relationship matrix. The non-zero components of this temporary vector determine the set of loop branches Ω={b11, b12, b19, b23, b27}. Then, considering the switch operability state, continuous load power supply requirements, and constraints for maintaining radial operation, the candidate disconnecting branch in Ω is determined to be the branch corresponding to S27. Subsequently, the branch rows corresponding to the candidate disconnecting branches are replaced with the branch rows corresponding to the closed branch L12, and a local update is performed based on the directional relationships of each branch in Ω to obtain the updated local branch current relationship matrix B. loc,1 and the updated local branch parameter vector Z loc,1 Through this process, connection reconfiguration and segmentation no longer trigger network-wide topology reconstruction, but instead complete incremental correction of the local topology representation within the initial influence domain.

[0038] For the injection change portion, the master station only updates the active and reactive injection parameters for the source and load nodes within D0 where injection changes occur. The active injection for PV2 nodes is corrected from 2.7MW to 0.9MW, and the reactive injection from 0.12Mvar to 0.05Mvar. For nodes within the influence domain that have not undergone injection changes, their pre-event injection parameters remain unchanged; for nodes outside the influence domain, their pre-event node voltage, branch power flow, and boundary exchange state remain unchanged to form a non-influence domain inherited state.

[0039] After completing the topology incremental update and injection incremental update, the master station updates the local branch current relationship based on the changes in switch states within the influence domain. It then combines the active and reactive power injection increments of the changed nodes with the directional branch current relationships corresponding to the boundary support branches to obtain the predicted active power exchange change and predicted reactive power exchange change at the boundary nodes. In this embodiment, the predicted active power exchange change at the boundary nodes is 1.32 MW, and the predicted reactive power exchange change is 0.18 Mvar. Simultaneously, the minimum electrical adjacency level from the event chain's action area to the master-distributor boundary node set is 1, and the propagation depth along the boundary coupling path is 3. The master station normalizes and weights the above factors to obtain the boundary penetration risk level value J. b =0.68. Because J bThe value is higher than the preset threshold of 0.60, therefore event chain C1 is determined to be a cross-domain event chain.

[0040] After C1 is identified as a cross-domain event chain, the master station performs a local incremental power flow simulation within the initial influence domain D0. Since D0 in this embodiment simultaneously contains both a local main grid coupling portion and a local distribution network radiation portion, the master station employs the sparse Newton method for the local main grid coupling portion and the forward-backward substitution method for the local distribution network radiation portion, exchanging local voltage and power information at the boundary nodes. After the first local incremental power flow simulation, the minimum node voltage within D0 is 0.948 pu, the maximum branch current utilization rate is 91.7%, the boundary active power mismatch is 1.12 MW, the boundary reactive power mismatch is 0.21 Mvar, and the boundary voltage deviation is 0.021 pu.

[0041] Since C1 is a cross-domain event chain and the boundary active power mismatch, boundary reactive power mismatch, and boundary voltage deviation do not simultaneously meet the stopping condition, the master station performs master-distributor boundary collaborative correction based on the boundary verification data M1. The master-distributor boundary collaborative correction includes: generating boundary equivalent injection correction and boundary voltage correction based on the boundary active power mismatch, boundary reactive power mismatch, and boundary voltage deviation between the local incremental power flow simulation results and the boundary measurement values ​​or the upper-level scheduling boundary state; updating the equivalent injection parameters and boundary node voltages of the boundary nodes within the influence domain based on these boundary equivalent injection correction and boundary voltage correction, and then re-executing the local incremental power flow simulation. After the first boundary collaborative correction, the boundary active power mismatch decreased to 0.46MW, the boundary reactive power mismatch decreased to 0.12Mvar, and the boundary voltage deviation decreased to 0.011pu, but the stopping condition still was not simultaneously met.

[0042] In this embodiment, the stopping condition is set as follows: the boundary active power mismatch is not greater than the first preset threshold τ. P The boundary reactive power mismatch is not greater than the second preset threshold τ. Q And the boundary voltage deviation is not greater than the third preset threshold τ V . Specifically, τ P Take 0.30MW, τ Q Take 0.15 Mvar, τ V Set the value to 0.010 pu. If any condition is not met, the stopping condition is considered not to have been met.

[0043] Since the stopping condition was not met after the first boundary collaborative correction, the master station expanded the influence domain outward according to the electrical adjacency layer. During the expansion, the master station selected the next-layer nodes and branches directly adjacent to the current influence domain based on the boundary support branches, the propagation path of the maximum boundary mismatch, and the node with the maximum boundary voltage deviation, and incorporated them into the influence domain, forming the expanded influence domain D1. After the expansion, 7 new nodes and 6 branches were added, including a section of the main branch upstream of F1 and two reactive power sensitive nodes in the middle and upper reaches of F2. After forming the expanded influence domain D1, the master station updated the local branch current relationship, the boundary coupling path set, and the injected parameters of the changing nodes, and recalculated the boundary penetration risk level. It then performed topology incremental update, injection incremental update, local incremental power flow simulation, and master-distributor boundary collaborative correction again. After the second round of solution, the boundary active power mismatch decreased to 0.22MW, the boundary reactive power mismatch decreased to 0.08Mvar, and the boundary voltage deviation decreased to 0.008pu, meeting the stopping condition. Therefore, the master-distributor collaborative power flow simulation results corresponding to this event chain were output. The final output includes the boundary node voltage, boundary switching power, voltage of each affected node, power flow of the affected branch, switch status, range of influence, and event chain type after the event.

[0044] After completing the above-mentioned complete event chain solution, at 14:12:09, the master station received the reactive power switching information of capacitor bank C23, forming a new event tuple E5. The event type of E5 is an injection event, the occurrence location is the upstream reactive power compensation device access point in F2, the affected device is C23, the occurrence time is 14:12:09.114, the change value is the reactive power compensation switching from 0 Mvar to 0.45 Mvar, and the data confidence level is 0.994. Since E5 and C1 meet the preset time window requirements, and C23 is located in the same propagation path after the connection reconfiguration, and acts together with C1 in the same main distribution boundary direction, the master station merges E5 into event chain C1. After the chain is merged, the master station re-establishes the corresponding main distribution joint influence diagram, expands the set of directly affected nodes from {L12, S27, PV2} to {L12, S27, PV2, C23}, and redetermines the initial influence domain. The updated initial influence domain contains 20 nodes and 19 branches. Boundary measurements are still used as boundary verification data and no event tuples are generated. Based on the updated topology increment, injection increment, and boundary support branch direction relationships, the master station obtains new predicted active power exchange changes and predicted reactive power exchange changes, and redetermines the boundary penetration risk level. Since the recalculated boundary penetration risk level is still higher than the preset boundary threshold, event chain C1 continues to maintain its cross-domain event chain attribute. The master station then performs iterative processes of local incremental power flow simulation, master-distribution boundary collaborative correction, and expansion of the influence domain according to the electrical adjacency layer, ultimately reducing the boundary active power mismatch to 0.22MW, the boundary reactive power mismatch to 0.09Mvar, and the boundary voltage deviation to 0.008pu, thus meeting the stopping conditions and outputting the updated master-distribution collaborative power flow simulation results. This process shows that when a new event tuple meets the time merging condition and satisfies the requirements of equipment topology reachability, causal continuity, power flow transmission continuity, and the direction of the joint action boundary, it can be merged into the existing event chain, directly changing the master-distribution joint influence diagram, influence domain, and subsequent solution path.

[0045] To further verify the effectiveness of this implementation method, the method of the present invention was compared with two control methods under the same operating scenario. The first control method solved E1, E2, E3, and E5 independently without constructing an event chain; the second control method performed a full network recalculation on all 58 distribution network nodes and boundary states after each event occurred. Table 1 shows the comparison results under the different control methods.

[0046] Table 1 Comparison of simulation results of master-slave coordinated incremental power flow under different event handling methods.

[0047] As shown in Table 1, under the same operating scenario and the same event input conditions, the cumulative number of nodes participating in the solution using the method of this invention is 47, significantly lower than the 86 of the single-event independent processing method and the 290 of the whole-network recalculation method. This indicates that the present invention can effectively narrow the solution range and reduce redundant calculations in irrelevant areas through event chain construction, main-distributor joint influence diagram establishment, and local update of the influence domain. Meanwhile, the influence domain expansion times and boundary coordination times of the method of this invention are 1 and 1, respectively, lower than the 3 times of the single-event independent processing method. This indicates that by determining the boundary penetration risk level, the propagation trend of the event to the main-distributor boundary can be identified in the initial stage, and the main-distributor boundary collaborative correction and influence domain expansion according to the electrical adjacency layer are triggered only when necessary. Furthermore, the boundary active power mismatch, boundary reactive power mismatch, and boundary voltage deviation finally obtained by the method of the present invention are 0.22MW, 0.08Mvar, and 0.008pu, respectively. These are on the same order of magnitude as the results of the full network recalculation method, which are 0.24MW, 0.09Mvar, and 0.009pu. This indicates that the present invention can maintain good simulation accuracy and engineering usability while reducing the solution scale and the number of solutions.

[0048] In addition, to illustrate the effectiveness of the closed-loop solution process within the method of the present invention, Table 2 shows the key state changes of event chain C1 at different solution stages.

[0049] Table 2 shows the key state changes in the event chain C1 solution process using the method of the present invention.

[0050] As shown in Table 2, after the first local incremental power flow simulation within the initial influence domain D0, the boundary active power mismatch, boundary reactive power mismatch, and boundary voltage deviation have not yet reached the stopping condition. After the primary and secondary boundary collaborative correction, all three indicators have decreased significantly, but the stopping condition is still not met simultaneously. Therefore, the influence domain is further expanded outward according to the electrical adjacency layer. After expanding the influence domain D1, the topology incremental update, injection incremental update, and local incremental power flow simulation are re-executed, so that the three boundary indicators simultaneously meet the stopping condition. This result shows that the closed-loop processing mechanism of "local incremental power flow simulation - primary and secondary boundary collaborative correction - expansion of influence domain according to electrical adjacency layer" adopted in this invention has a clear convergence path and good engineering applicability.

[0051] At 14:12:31, after this embodiment concludes, the master station receives the remote small load disturbance information from F4, forming event tuple E6. Since this event does not meet the preset time window requirement with C1, and its area of ​​action has no direct topological reachability to the transfer paths of F1 and F2, nor does it share a common boundary direction, the master station does not merge it into C1, but treats it as an independent event chain. In this way, the system can simultaneously support the parallel chain solution of associated events and the independent solution of unassociated events within the same operating cycle, ensuring the adaptability of the master-distributor coordinated incremental power flow simulation to continuously arriving, multi-source concurrent discrete events.

[0052] Example 2 This embodiment illustrates the execution of the method of the present invention in a local event chain scenario when discrete events do not exhibit a clear boundary penetration trend. The operating conditions are the same as in Embodiment 1, except that a single load disturbance at the F4 remote end is selected as the input event to illustrate the solution process of the local event chain within the event-driven primary-distributor coordinated incremental power flow simulation framework.

[0053] At 15:03:16, the master station received power fluctuation information from an industrial load node at the end of F4, forming event tuple E7. Event type E7 is an injection event, occurring at the remote load node of F4, with the acting device being the power acquisition terminal of that node. The occurrence time is 15:03:16.447, and the changes are: active load increasing from 0.62MW to 0.74MW and reactive load increasing from 0.21Mvar to 0.26Mvar, with a data confidence level of 0.993. Since there are no other event tuples within the preset time window that satisfy the requirements of device topology reachability, causal continuity, power flow continuity, and common action boundary direction, the master station directly processes E7 as an independent event chain C3.

[0054] After event chain C3 is formed, the master station establishes a corresponding main-distribution joint impact diagram. The nodes of the joint impact diagram include main grid nodes, distribution network nodes, boundary nodes, load nodes, and adjacent switch operation nodes; the edges of the joint impact diagram include transmission line edges, distribution branch edges, main-distribution boundary coupling edges, and event causal edges. The master station determines the initial influence domain D3 based on the directly acting load node of E7, the adjacent branches connected to that node, and the adjacent nodes and branches along the power flow direction. D3 contains 6 nodes and 5 branches, but does not include other feeder areas unrelated to this event.

[0055] After obtaining the initial influence domain D3, the master station only performs incremental injection updates on nodes within D3 where injection changes have occurred, updating the active power injection and reactive power injection of these load nodes to 0.74MW and 0.26Mvar, respectively. For nodes within D3 where no injection changes have occurred, the pre-event injection parameters remain unchanged. For nodes outside D3 that are not in the influence domain, the pre-event node voltage, branch power flow, and boundary exchange state remain unchanged. Then, based on the updated injection parameters of the changed nodes, the direction relationship of the boundary support branches, and the minimum electrical adjacency level from the current influence domain to the set of main and distribution boundary nodes, the master station obtains a predicted active power exchange change of 0.04MW and a predicted reactive power exchange change of 0.02Mvar at the boundary nodes, with a relatively small propagation depth along the boundary coupling path. After normalized weighting, the boundary penetration risk level value J_b=0.11, which is lower than the preset boundary threshold of 0.60. Therefore, event chain C3 is determined to be a local event chain.

[0056] After C3 was identified as a local event chain, the master station only performed local incremental power flow simulation within D3. The results showed that the voltage of the most unfavorable node in D3 decreased from 0.972 pu to 0.968 pu, the local branch current utilization increased from 63.5% to 66.1%, the boundary active power mismatch was 0.04 MW, the boundary reactive power mismatch was 0.02 Mvar, and the boundary voltage deviation was 0.002 pu. Since these three indicators met the first, second, and third preset thresholds respectively, the master station directly output the power flow simulation results of the local event chain without triggering master-distributor boundary collaborative correction and influence domain expansion.

[0057] This embodiment illustrates that, in local operational disturbance scenarios without boundary penetration trends, the method of the present invention can solve the operational state after an event through independent event chains, local primary and secondary joint influence diagrams, local injection incremental updates, and local incremental power flow simulation, thereby enabling local event chains and cross-domain event chains to be processed uniformly within the same method framework.

[0058] In another optional scenario, if any of the boundary active power mismatch, boundary reactive power mismatch, and boundary voltage deviation corresponding to the local incremental power flow simulation results of the local event chain do not meet the corresponding threshold, the master station will transfer the local event chain into the master-distributor boundary collaborative correction process. If necessary, after expanding the influence domain outward according to the electrical adjacent layer, the local incremental power flow simulation and master-distributor boundary collaborative correction will be repeatedly executed until the stopping condition is met and the results are output.

[0059] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0060] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An event-driven based master-slave coordination incremental power flow simulation method, characterized in that, include: The system acquires the transmission network model, distribution network model, main and distribution boundary node set, and node voltage, branch power flow, and boundary exchange status before the event occurs; it receives discrete events and generates event tuples, which are then grouped into event chains according to a preset time window and equipment topology reachability; it establishes a main and distribution joint impact diagram based on the event chains, and determines the initial influence domain based on the directly affected nodes of the event chains, branches affected by topology changes, and adjacent nodes and branches connected along the power flow propagation direction; and it calculates the boundary exchange power change based on the minimum electrical adjacency level from the initial influence domain to the main and distribution boundary node set, the boundary coupling path propagation depth, and the predicted boundary exchange power change obtained from the topology increment and injection increment within the influence domain. The boundary penetration risk level is determined, and the event chain is divided into local event chains or cross-domain event chains. Topology incremental updates and injection incremental updates are performed only within the influence domain, while maintaining the pre-event state of non-influence domains for local incremental power flow simulation. When any of the boundary active power mismatch, boundary reactive power mismatch, and boundary voltage deviation does not meet the corresponding threshold, primary and secondary boundary collaborative correction is performed. If the stopping condition is still not met after correction, the influence domain is expanded outward according to the electrical adjacency layer, and the boundary penetration risk level determination, incremental update, local incremental power flow simulation, and primary and secondary boundary collaborative correction are repeated until the stopping condition is met, and then the primary and secondary collaborative power flow simulation results after the event are output.

2. The event-driven master-slave coordinated incremental power flow simulation method according to claim 1, characterized in that, The event tuple includes event type, occurrence location, affected device, occurrence time, change value, and data confidence level. The event type includes topology events and injection events. The topology events include tie switch closing events, sectionalizing switch opening events, sectionalizing switch closing events, and feeder transfer events. The injection events include distributed generation active power output change events, distributed generation reactive power output change events, load change events, and reactive power compensation device switching events. Boundary measurement values ​​are only used for verification of local incremental power flow simulation results and are not used as constituent elements of the event chain.

3. The event-driven master-slave coordinated incremental power flow simulation method according to claim 1, wherein, The process of grouping event elements into event chains based on a preset time window and device topology reachability includes: grouping event elements whose occurrence time difference does not exceed the preset time window and which act on the same feeder, the same branch switching channel, or adjacent electrical equipment into the same candidate event chain; and then determining the candidate event chain as the target event chain based on the causal continuity, power flow transmission continuity, and common action boundary direction between the event elements.

4. The event-driven master-slave coordinated incremental power flow simulation method according to claim 1, wherein, Before constructing the initial influence domain, a main-distribution joint influence diagram is established. The nodes of the main-distribution joint influence diagram include main grid nodes, distribution network nodes, boundary nodes, switch action nodes, and distributed power generation nodes or load nodes. The edges of the main-distribution joint influence diagram include transmission line edges, distribution branch edges, main-distribution boundary coupling edges, and event causal edges. The initial influence domain is determined based on the directly affected nodes of the event chain, the branches affected by topology changes, and the adjacent nodes and branches connected to the directly affected nodes along the power flow transmission direction.

5. The event-driven master-slave coordinated incremental power flow simulation method according to claim 1, wherein, The boundary penetration risk level is determined by a combination of the following factors: the minimum electrical adjacency level from the event chain's action area to the set of primary and secondary boundary nodes, the propagation depth of the event chain along the boundary coupling path, the predicted change in boundary exchange power caused by topological increments within the affected domain, and the predicted change in boundary exchange power caused by injection increments within the affected domain. When the boundary penetration risk level is lower than a preset boundary threshold, the event chain is determined to be a local event chain; otherwise, it is determined to be a cross-domain event chain.

6. The event-driven master-slave coordinated incremental power flow simulation method according to claim 5, characterized in that, The predicted boundary exchange power change is obtained by updating the local branch current relationship based on the change of switch state in the influence domain, and combining the active power injection increment and reactive power injection increment of the changed node with the directional branch current relationship corresponding to the boundary support branch to obtain the predicted results of the active power exchange change and reactive power exchange change at the boundary node.

7. The event-driven master-slave coordinated incremental power flow simulation method according to claim 1, wherein, The primary-secondary boundary collaborative correction includes: generating boundary equivalent injection correction and boundary voltage correction based on the boundary active power mismatch, boundary reactive power mismatch, and boundary voltage deviation between the local incremental power flow simulation results and the boundary measurement values ​​or the upper-level scheduling boundary state; and updating the equivalent injection parameters and boundary node voltages of the boundary nodes within the influence domain based on the boundary equivalent injection correction and boundary voltage correction, and then re-executing the local incremental power flow simulation.

8. The event-driven master-distributor cooperative incremental power flow simulation method according to claim 1, characterized in that, The incremental injection update includes: updating the active power injection parameters and reactive power injection parameters of source nodes and load nodes that have undergone injection changes within the influence domain; keeping the pre-event injection parameters unchanged for nodes that have not undergone injection changes within the influence domain; and keeping the pre-event node voltage, branch power flow, and boundary exchange status unchanged for nodes outside the influence domain.

9. The event-driven master-slave coordinated incremental power flow simulation method according to claim 1, wherein, The stopping conditions are that the boundary active power mismatch is not greater than the first preset threshold, the boundary reactive power mismatch is not greater than the second preset threshold, and the boundary voltage deviation is not greater than the third preset threshold. If any condition is not met, it is determined that the stopping condition has not been met.

10. The event-driven master-slave coordinated incremental power flow simulation method according to claim 1, wherein, The expansion of the influence domain outward from the electrical adjacency layer includes: incorporating the next layer nodes and branches directly adjacent to the boundary support branch, the maximum mismatch propagation path, or the node with the maximum boundary voltage deviation outside the current influence domain into the influence domain; updating the local branch current relationship, the boundary coupling path set, and the variable node injection parameters within the expanded influence domain; and repeatedly performing local incremental power flow simulation and main-distribution boundary collaborative correction.

Images