Search

CN-121480989-B - Power distribution network toughness improving method based on data fusion algorithm

CN121480989BCN 121480989 BCN121480989 BCN 121480989BCN-121480989-B

Abstract

The invention belongs to the technical field of power distribution networks, and provides a power distribution network toughness improving method based on a data fusion algorithm, which comprises the steps of collecting and fusing multi-dimensional data of the power distribution network, identifying regional-level toughness blocks with similar toughness characteristics by using a clustering algorithm, and identifying system-level toughness key nodes with the greatest influence on the overall survivability of a system by combining a complex network theory and fault simulation to form a toughness region and key node set; aiming at the toughness region and the key node set, evaluating the node level, region level and system level toughness index values corresponding to different distributed resource space allocation schemes under a preset extreme scene, and calculating the toughness increment by comparing whether the toughness increment exists or not. An optimization model integrating the verified causal relationship and multi-objective coordination is established, and a scientific resource allocation strategy which ensures the maximization of the overall toughness of the system and gives consideration to the balance of the toughness among areas is output, so that the precision, reliability and coordination promotion of the toughness of the power distribution network are realized.

Inventors

  • YU HAI

Assignees

  • 安极能新能源发展有限公司

Dates

Publication Date
20260512
Application Date
20260105

Claims (7)

  1. 1. A power distribution network toughness improving method based on a data fusion algorithm is characterized by comprising the following steps: acquiring and fusing multidimensional data of a power distribution network, identifying regional-level toughness blocks with similar toughness characteristics by using a clustering algorithm, and identifying system-level toughness key nodes with the greatest influence on the overall survivability of the system by combining a complex network theory and fault simulation to form a toughness region and key node set; Aiming at the toughness region and the key node set, evaluating node level, region level and system level toughness index values corresponding to different distributed resource space configuration schemes under a preset extreme scene, calculating toughness increment by comparing whether the toughness increment exists or not, and identifying pseudo-related candidate hot spots influenced by hidden variables; based on the toughness index value, comparing and analyzing to improve the competition or conflict between the space and the capacity of the resources required by the regional toughness balance and the resources required by the toughness key nodes of the strengthening system, judging whether the contradiction exists or not; The verified resource-toughness mapping table includes: a resource contradiction region, namely spatial position information of direct resource competition exists between key nodes for improving the toughness balance of the region and strengthening the toughness of the system; The high-priority synergistic object is nodes and areas with real and obvious forward causal effect on system toughness improvement through inverse fact simulation verification and resource investment; pseudo-relevant or low-efficiency hot spots, namely nodes and areas which are weakly or invalidily related to the resource investment and toughness improvement due to third variable interference or improper original configuration; if the contradiction exists, constructing a multi-objective optimization model integrating the regional balance requirement with the aim of maximizing the system-level toughness, and solving an optimal resource allocation strategy of output coordination regional balance and key protection; The process for constructing the multi-objective optimization model fusing the regional balance requirement comprises the following steps: defining decision variables, namely planning newly-increased installed capacity of the distributed power supply and the schedulable capacity value of the electric vehicle by using each power grid node as decision variables; setting a core objective function, wherein a main objective is to maximize an expected improvement value of a system-level toughness index, and a subordinate objective is to minimize the degree of difference of final toughness levels of all area-level toughness blocks; Constructing constraint conditions, wherein the constraint conditions comprise total resource capacity constraint, upper and lower limit constraint of each node resource, promotion constraint of a high-priority synergistic object based on a verified resource-toughness mapping relation table, avoidance constraint of pseudo-correlation or low-efficiency hot spots, coordination constraint aiming at resource contradiction areas and tide and voltage constraint for guaranteeing safe operation of a power grid; the process for solving the optimal resource allocation strategy of the balance and key protection of the output coordination area is as follows: Solving the model by adopting an intelligent optimization algorithm, and obtaining a group of Pareto optimal solution sets representing different trade-offs after convergence; the Pareto optimal solution set is visually presented by taking the system-level toughness improvement value and the regional toughness balance index as dimensions; And selecting a final scheme from the Pareto optimal solution set according to a preset decision rule, and expressing the final scheme as an optimal distributed resource space configuration strategy containing specific resource types, capacities and expected toughness benefits of each node.
  2. 2. The method for improving toughness of a power distribution network based on a data fusion algorithm according to claim 1, wherein the process of identifying regional-level toughness blocks with similar toughness characteristics by using a clustering algorithm is as follows: Calculating a multidimensional toughness feature vector for each power grid node, wherein the feature dimensions comprise structural vulnerability, historical reliability, risk exposure and resource supporting potential; And clustering the normalized feature vectors, and defining a node set which belongs to the same cluster and is continuous or adjacent in space distribution as a regional-level toughness block.
  3. 3. The method for improving the toughness of the power distribution network based on the data fusion algorithm of claim 1, wherein the process of identifying the system-level toughness key node with the greatest influence on the overall survivability of the system is as follows: Performing complex network topology analysis, and calculating topology centrality indexes of each node, including electrical betweenness centrality and proximity centrality; constructing and executing preset extreme disturbance scene simulation, and quantitatively evaluating influence severity indexes of each node failure, wherein the influence severity indexes comprise a load loss influence range, a recovery path blocking degree and a cascading failure triggering risk; the topology center index and the influence severity index are fused, the nodes are comprehensively ordered based on a TOPSIS method, and the nodes with the top order are determined to be system-level toughness key nodes by setting a ranking proportion threshold value.
  4. 4. The method for improving the toughness of the power distribution network based on the data fusion algorithm of claim 1, wherein the process of evaluating the toughness index values of the node level, the area level and the system level corresponding to different distributed resource space allocation schemes in a preset extreme scene is as follows: Defining a node level, a regional level and a system level toughness quantization index system; Based on the toughness region and the key node set, generating a plurality of distributed resource space configuration schemes with different resource distribution trends; the distributed resource space allocation scheme comprises a key node reinforcement type, a weak area reinforcement type and a uniform dispersion type; aiming at each distributed resource space configuration scheme, executing time sequence simulation by combining with a preset extreme scene set; Calculating toughness index values of each node, each regional level toughness block and the whole system under each scheme-scene combination according to the simulation output data; and comparing the toughness index values of the schemes with the standard scheme, and calculating to obtain the increment of each level of toughness index.
  5. 5. The method for improving the toughness of the power distribution network based on the data fusion algorithm of claim 1, wherein the method comprises the following steps of: Identifying and recording grid nodes of which the allocated resource capacity in the resource allocation scheme is higher than the scheme average level but at least one of the node level, the affiliated regional level toughness block level or the system level toughness increment is lower than the scheme average increment level through high-projection low-efficiency screening; Identifying and recording grid nodes or regional-level toughness blocks with the toughness increment value higher than the average level of the whole grid nodes or regions among different polar-end disturbance scenes through scene-dependent screening; And sorting and summarizing the results of the high-projection low-efficiency screening and the scene-dependent screening to form a pseudo-relevant candidate hot spot list.
  6. 6. The method for improving the toughness of the power distribution network based on the data fusion algorithm of claim 4, wherein the process for judging whether the contradiction phenomenon exists is as follows: forming a region balance enhanced resource demand graph and a key node enhanced resource demand graph respectively based on simulation results of the weak region enhanced scheme and the key node enhanced scheme; comparing the two resource demand graphs, and identifying nodes or areas which are overlapped or adjacent in physical space and have opposite resource demand directions; under the constraint of the total newly increased resource budget of the system, judging whether meeting the requirement of one party can not fully meet the requirement of the other party, if so, judging that the resource contradiction exists.
  7. 7. The method for improving toughness of a power distribution network based on a data fusion algorithm according to claim 5, wherein the process of distinguishing the true correlation from the false correlation is as follows: for objects in the pseudo-relevant candidate hotspot list; selecting potential third variables, layering all nodes according to values, and re-analyzing the relation between the resource investment and the toughness increment of the object in each layer to judge whether abnormal association is dominated by the corresponding variables; The anti-facts simulation verification, namely, designing a simulation scene aiming at an object to be analyzed, and virtually reconfiguring resources planned to be input to a control node which has similar topology or load characteristics but is not marked as a hot spot; And judging the real causal effect of the input of the corresponding object resources by comparing the incremental changes of the system-level toughness indexes before and after the resource reconfiguration, and further dividing the real causal effect into high-priority synergistic objects or pseudo-related hot spots.

Description

Power distribution network toughness improving method based on data fusion algorithm Technical Field The invention belongs to the technical field of power distribution networks, and particularly relates to a power distribution network toughness improving method based on a data fusion algorithm. Background The toughness of the power distribution network refers to the capability of preventing, resisting, quickly recovering and adapting to the power grid after the power grid suffers from extreme disturbance (such as extreme weather and network attack), emphasizes that the system is not only required to be knocked down (reliability), but also required to be quickly standing up after being knocked down, ensures continuous power supply to the maximum extent, maintains key functions, and has the fundamental transition of the form and the operation paradigm of the power distribution network along with the wide access of a high-proportion distributed power supply and an electric automobile, wherein the distributed resource brings considerable flexibility and local support potential, but the strong randomness and space-time coupling characteristic of the source-charge double sides also lead the complexity of the system to be greatly increased, and the toughness characterization of the power distribution network presents new characteristics of cross-scale and nonlinearity; However, the traditional planning and analysis method is mostly based on a deterministic static model and a single data dimension, the dynamic evolution process is difficult to describe, the inherent logic tends to be hidden with local enhancement to inevitably linearly promote global simple assumption, which leads to possible local optimal damage global toughness paradox in practice, for example, resources are excessively intensively deployed in few nodes, although strong toughness islands can be formed, supporting paths for weak areas can be cut off due to fault isolation, the whole collaborative recovery capability of a system is weakened, the deeper contradiction is that the improvement of toughness faces two difficult choices, namely, regional toughness requires resource dispersion configuration to realize equalization and avoid systematic short plates, and the system toughness also needs resources to incline to topological and functional key nodes to inhibit linkage faults, the traditional method lacks quantitative cognition of nonlinear mapping relation among multi-level toughness indexes, and tends to be difficult to experience or single-target optimization in the face of the contradiction, so that real extreme risks cannot be absorbed; Therefore, the invention provides a power distribution network toughness improving method based on a data fusion algorithm. Disclosure of Invention In order to overcome the deficiencies of the prior art, at least one technical problem presented in the background art is solved. The technical scheme adopted for solving the technical problems is that the method for improving the toughness of the power distribution network based on the data fusion algorithm comprises the following steps: acquiring and fusing multidimensional data of a power distribution network, identifying regional-level toughness blocks with similar toughness characteristics by using a clustering algorithm, and identifying system-level toughness key nodes with the greatest influence on the overall survivability of the system by combining a complex network theory and fault simulation to form a toughness region and key node set; Aiming at the toughness region and the key node set, evaluating node level, region level and system level toughness index values corresponding to different distributed resource space configuration schemes under a preset extreme scene, calculating toughness increment by comparing whether the toughness increment exists or not, and identifying pseudo-related candidate hot spots possibly influenced by hidden variables; based on the toughness index value, comparing and analyzing to improve the competition or conflict between the space and the capacity of the resources required by the regional toughness balance and the resources required by the toughness key nodes of the strengthening system, judging whether the contradiction exists or not; if the contradiction exists, a multi-objective optimization model fused with the regional balance requirement is constructed with the aim of maximizing the system-level toughness, and an optimal resource allocation strategy of the output coordination region balance and the key protection is solved. The beneficial effects of the invention are as follows: According to the invention, through constructing a logic closed loop of identification, evaluation, diagnosis and optimization, the two core contradictions of local optimal damage global and area balance and key protection target conflict faced by the toughness promotion of the power distribution network under the condition of accessing a hig