CN-122021351-A - Substation intelligent configuration optimization method and system based on modularized design

CN122021351ACN 122021351 ACN122021351 ACN 122021351ACN-122021351-A

Abstract

The application provides a substation intelligent configuration optimization method and system based on modular design, and relates to the technical field of configuration optimization, wherein the method comprises the steps of obtaining an initial module set of a substation; the method comprises the steps of establishing a three-dimensional configuration model based on an initial module set, generating a module dependency tree according to the three-dimensional configuration model, obtaining an initial module configuration scheme based on the module dependency tree, generating a first generation population containing reference individuals and variant individuals based on the initial module configuration scheme and a genetic algorithm, executing a composite verification condition to obtain an effective population, generating a next generation population according to the effective population, and executing multiple iterations to obtain an optimal configuration scheme. Through changing the improvement of the genetic algorithm into an intelligent optimizing method driven by a computer, the compactness and the rationality of the layout of the substation module are greatly improved.

Inventors

CHEN WEIQI
CHEN ZHENYU
ZHANG KA
ZHOU TAO
LIU GUOSHENG
ZHANG JINSHAN
ZHU LEI
FU CHUNYANG
Yi Tangxiang

Assignees

湘能楚天电力装备股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260413

Claims (8)

1. The intelligent configuration optimization method for the transformer substation based on the modularized design is characterized by comprising the following steps of: Step 1, acquiring an initial module set of a transformer substation; Step 2, constructing a three-dimensional configuration model based on the initial module set; generating a module dependency tree according to the three-dimensional configuration model, and obtaining an initial module configuration scheme based on the module dependency tree, wherein the modules are used as nodes, and the nodes are connected through electrical directed edges to form the module dependency tree; and 4, generating a first generation population containing reference individuals and variant individuals based on the initial module configuration scheme and a genetic algorithm, executing a composite verification condition to obtain an effective population, generating a next generation population according to the effective population, and executing multiple iterations to obtain an optimal configuration scheme.
2. The substation intelligent configuration optimization method based on modular design according to claim 1, wherein constructing a three-dimensional configuration model based on the initial module set comprises the following steps: Step 21, acquiring a three-dimensional coordinate point of each module and an electric adjacent matrix among the modules according to the initial module set; Step 22, constructing a three-dimensional coordinate system, and dividing the three-dimensional coordinate system into an operation area and a forbidden area based on the three-dimensional coordinate system, wherein the operation area is modeled as a union set formed by a plurality of convex polyhedrons; Step 23, if the three-dimensional coordinate point of the current module is positioned in the interior or the surface of the convex polyhedron in the union set, judging that the current module is positioned in an operation area, otherwise, judging that the current module is positioned in a forbidden area; and step 24, generating a three-dimensional configuration model according to the three-dimensional coordinate points, the electric adjacency matrix, the operation area and the forbidden area.
3. The substation intelligent configuration optimization method based on the modular design according to claim 2 is characterized in that a first ray is emitted to a convex polyhedron of the union set based on the current three-dimensional coordinate point, the number of intersection points of the first ray and the surface of the convex polyhedron is determined, if the number of intersection points is odd, the current three-dimensional coordinate point is determined to be positioned in the convex polyhedron of the union set, and if the number of intersection points is even, the current three-dimensional coordinate point is determined to be positioned outside the convex polyhedron of the union set.
4. The intelligent configuration optimization method for a transformer substation based on modular design according to claim 3, wherein the initial module configuration scheme comprises three-dimensional coordinate points, rotation angles and model parameters of all modules.
5. The intelligent configuration optimization method of a transformer substation based on modular design according to claim 4, wherein generating a first generation population including reference individuals and variant individuals based on the initial module configuration scheme and genetic algorithm, performing a composite verification condition to obtain an effective population, generating a next generation population according to the effective population, and performing multiple iterations to obtain an optimal configuration scheme, comprises the following steps: Step 41, constructing a reference gene sequence according to three-dimensional coordinate points, rotation angles and model parameters of all modules in the initial module configuration scheme; step 42, assigning the reference gene sequence to a first individual in the first generation population as a reference individual; Step 43, based on n-1 individuals remaining in the first generation population, performing initialization operation according to the three-dimensional coordinate points, the rotation angles and the model parameters to obtain a first parameter set and form n-1 variant individuals, and generating the first generation population according to the reference individuals and the variant individuals; step 44, based on the reference individuals and variant individuals of the first generation population, performing a composite verification condition, eliminating the reference individuals and/or variant individuals failing to verify, and reserving the reference individuals and/or variant individuals successful in verification as an effective population; Step 45, calculating fitness values of all the reference individuals and/or variant individuals in the effective population, sorting all the reference individuals and/or variant individuals in the effective population based on the fitness values, taking M reference individuals and/or variant individuals before sorting as candidate individuals, wherein M is greater than or equal to 4, and executing gene sequence crossing and variant operations based on the candidate individuals to generate a next generation population; Step 46, repeating steps 43-45 until the iteration number reaches a preset threshold or the change rate of the fitness value of the continuous five-generation population based on the first-generation population is smaller than the preset convergence precision, and generating an optimal configuration scheme based on the individual with the highest fitness value.
6. The substation intelligent configuration optimization method based on modular design according to claim 5, wherein based on n-1 individuals remaining in the first generation population, performing initialization operation according to the three-dimensional coordinate point, the rotation angle and the model parameters to obtain a first parameter set and form n-1 variant individuals, and generating the first generation population according to the reference individuals and the variant individuals, comprising the following steps: step 431, constructing an initial random offset vector, and generating a first coordinate point based on the three-dimensional coordinate points of each module and the initial random offset vector; step 432, executing the judgment of the operation area and the forbidden area based on the first coordinate point, if the first coordinate point is positioned in the operation area, taking the first coordinate point as a new coordinate point of the module, otherwise, executing step 433 to carry out boundary correction; Step 433, presetting a maximum retry number k, and in a cycle of k=1 to k, re-executing steps 431-432 to generate an initial random offset vector and a first coordinate point until the first coordinate point is located in the operation area, and ending with the first coordinate point as a new coordinate point of the module; Step 434, if the first coordinate point is still in the forbidden region in step 433, performing ray casting calculation, namely drawing a second ray to the union set along the direction of an initial random offset vector by taking the three-dimensional coordinate point as a starting point, calculating an effective intersection point of the second ray and the union set, generating a new coordinate point according to the three-dimensional coordinate point and the effective intersection point, and if no effective intersection point exists, reversing the initial random offset vector, and re-performing ray casting calculation to obtain the new coordinate point; Step 435, presetting a maximum rotation threshold, generating an angle constraint interval based on the maximum rotation threshold, generating random floating point numbers by adopting a Meisson rotation algorithm based on the angle constraint interval, and generating a new rotation angle based on the random floating point numbers and the rotation angle; Step 436, repeatedly executing steps 431-435 to generate new coordinate points, new rotation angles and replacement models of all modules in n-1 variant individuals, generating complete gene sequences of each variant individual based on the new coordinate points, the new rotation angles and the replacement models of the modules, and combining the reference individual with the n-1 variant individuals to form a first generation population containing n individuals.
7. The substation intelligent configuration optimization method based on modular design according to claim 6, wherein, based on the reference individuals and variant individuals of the first generation population, performing a composite verification condition, eliminating the reference individuals and/or variant individuals failing to be verified, and reserving the reference individuals and/or variant individuals failing to be verified as an effective population, comprising the following steps: step 441, obtaining three-dimensional entity models of all modules based on the reference gene sequence and the complete gene sequence; Step 442, grouping based on all modules to obtain grouping modules, determining boolean intersections of three-dimensional entity models of all grouping modules in the reference individual and/or variant individual, and determining boolean intersection volumes based on the boolean intersections; 443, if the boolean intersection volume of the three-dimensional solid models of any grouping module is greater than 0, determining that the current reference individual or variant individual has geometric interference failure, and eliminating the current reference individual or variant individual; Step 444, acquiring adjacent module pairs of the reference individual and/or the variant individual according to the module dependency tree, wherein the adjacent module pairs comprise an upstream node and a downstream node, determining primary labels, secondary labels and tertiary labels of the upstream node and the downstream node, judging that the current reference individual or variant individual passes the verification if the following conditions are met, otherwise, rejecting the current reference individual or variant individual: the method comprises the following steps that 1, a primary label of an upstream node is consistent with a primary label of a downstream node; the condition 2 is that the secondary label of the upstream node is the same as the secondary label of the downstream node, or the secondary label of the downstream node belongs to a first subset defined by the secondary label of the upstream node; determining a third-level first sub-tag and a third-level second sub-tag according to the third-level tag, wherein the third-level first sub-tag of the upstream node is identical to the third-level first sub-tag of the downstream node, and the third-level second sub-tag of the downstream node is identical to the third-level second sub-tag of the upstream node or the third-level second sub-tag of the downstream node belongs to a second subset of the third-level second sub-tag of the upstream node; step 445 defines the reference individual and/or variant individuals that were successfully verified as an effective population.
8. A substation intelligent configuration optimization system based on modular design, configured to execute the substation intelligent configuration optimization method based on modular design as set forth in any one of claims 1 to 7, and characterized by comprising the following modules: The acquisition module is used for acquiring an initial module set of the transformer substation; the construction module is used for constructing a three-dimensional configuration model based on the initial module set; The initial configuration module is used for generating a module dependency tree according to the three-dimensional configuration model and obtaining an initial module configuration scheme based on the module dependency tree; And the optimizing module is used for generating a first generation population containing reference individuals and variant individuals based on the initial module configuration scheme and the genetic algorithm, executing the composite verification condition to obtain an effective population, generating a next generation population according to the effective population, and executing multiple iterations to obtain the optimizing configuration scheme.

Description

Substation intelligent configuration optimization method and system based on modularized design Technical Field The application relates to the technical field of configuration optimization, in particular to a substation intelligent configuration optimization method and system based on modular design. Background With the continuous promotion of smart grid construction, a modular design is used as a standardized and intensive engineering methodology, and gradually becomes a mainstream trend of substation construction. The modular design not only refers to dividing the transformer substation into a plurality of physical units with independent functions, but also emphasizes that the physical units are converted into digital modules which can be identified by a computer, namely, a digital twin body comprising the geometric dimension, the electrical interface and the functional attribute of the modules is established in the initial stage of design. The current technical means can generate an initial configuration scheme containing specific three-dimensional coordinate points, rotation angles and model parameters of each module through three-dimensional modeling and other modes based on an initial module set of the transformer substation. This initial solution provides a basic data framework for the physical layout of the substation, which is the starting point for the subsequent design work. However, from an engineering practice perspective, there are significant limitations to existing initial configuration scheme generation methods. The existing method is usually focused on preliminary placement and static definition of modules, and lacks a set of dynamic mechanisms capable of taking the initial scheme as input and performing deep iterative optimization on the basis. In particular, it is difficult in the prior art to effectively address how to systematically explore and generate a variety of configuration possibilities that are more compact in physical space layout and more reasonable in electrical connection relationships based on a viable initial solution. Because of the lack of a complete process for intelligent evolution, screening and optimization of the initial scheme, designers often need to rely on a large number of manual trial and error and repeated adjustment to obtain a configuration scheme which finally meets all complex constraint conditions. This process is not only inefficient, but it is also difficult to guarantee the optimality of the final solution in a global range, especially in the face of large substation projects with numerous modules and complex connections, which is a drawback of the prior art. Therefore, how to construct a set of method capable of automatically performing multi-generation iteration and optimization through an intelligent algorithm based on an initial configuration scheme so as to efficiently generate a high-quality transformer substation configuration scheme becomes a technical problem to be solved in the field. Disclosure of Invention The present application aims to solve at least one of the technical problems in the related art to some extent. To achieve the above objective, the embodiment of the present application provides a substation intelligent configuration optimization method and system based on modular design, including the following steps: Step 1, acquiring an initial module set of a transformer substation; Step 2, constructing a three-dimensional configuration model based on the initial module set; generating a module dependency tree according to the three-dimensional configuration model, and obtaining an initial module configuration scheme based on the module dependency tree, wherein the modules are used as nodes, and the nodes are connected through electrical directed edges to form the module dependency tree; and 4, generating a first generation population containing reference individuals and variant individuals based on the initial module configuration scheme and a genetic algorithm, executing a composite verification condition to obtain an effective population, generating a next generation population according to the effective population, and executing multiple iterations to obtain an optimal configuration scheme. Further, constructing a three-dimensional configuration model based on the initial set of modules, comprising the steps of: Step 21, acquiring a three-dimensional coordinate point of each module and an electric adjacent matrix among the modules according to the initial module set; Step 22, constructing a three-dimensional coordinate system, and dividing the three-dimensional coordinate system into an operation area and a forbidden area based on the three-dimensional coordinate system, wherein the operation area is modeled as a union set formed by a plurality of convex polyhedrons; Step 23, if the three-dimensional coordinate point of the current module is positioned in the interior or the surface of the convex polyhedron in the union set, jud