Method and system for selecting and matching data organization structure of power system model

Abstract

The invention provides a data organization structure matching method and system for power system models, and the method comprises the steps: obtaining a plurality of power system models in a power system and a plurality of selectable data organization structures, and constructing a plurality of control variables according to the power system models and the selectable data organization structures; constructing constraint conditions of control variables and a target function of data organization structure selection and matching based on performance requirements of each power system model and each data organization structure; according to each control variable, the constraint condition and the objective function, a preset genetic algorithm is used for solving, and a solved power system data organization structure selection scheme is obtained; and according to the type of each power model, merging each power system model and each data organization structure in the solved power system data organization structure matching scheme to obtain a final power system data organization structure matching scheme. According to the system, the matching efficiency of the power system model data organization structure scheme is improved.

Description

Technical Field

[0001] The present application relates to the technical field of power system model data integration, and in particular to a method and system for selecting a data organization structure for a power system model. Background Art

[0002] In power grid control services, different data organization methods can lead to different extraction efficiencies for the same business retrieval requirements, impacting overall business response time. Existing research on the integration and organization of multi-source heterogeneous data primarily considers the organizational structure of the data source, conducting research on multi-source data semantic fusion and integrated data query to create a universal data set with complete data, unified semantics, and low redundancy. However, this research lacks consideration of the performance requirements of specific business operations for data extraction. Directly applying universal methods to build a data platform will make it difficult to meet the real-time requirements of power grid control services.

[0003] To meet the data retrieval performance requirements of specific businesses, the power industry primarily relies on the engineering experience of technicians to select the appropriate power system model data organization structure. This approach also lacks relevant qualitative analysis and evaluation methods, resulting in significant subjectivity and difficulty in widespread application. Therefore, it is crucial to research data organization optimization methods for the multi-business requirements of power grid regulation. By selecting appropriate data organization structures for each attribute and relationship of the power system model to organize and store data, a large number of power grid regulation businesses can meet performance requirements when acquiring the required model data. Abstract modeling of this data organization scheme and computational analysis of related data redundancy can provide a data basis for the design of related storage architectures and the planning and selection of storage media, thereby avoiding the high costs of data synchronization and management. Summary of the Invention

[0004] In response to the above technical problems, the present application provides a method for selecting the data organization structure of a power system model to improve the business performance of each power system model and the selection efficiency of the data organization structure.

[0005] In a first aspect, an embodiment of the present application provides a method for selecting a data organization structure of a power system model, comprising:

[0006] Acquire several power system models and several optional data organization structures in the power system;

[0007] Constructing a plurality of control variables according to each of the power system models and each of the data organization structures, wherein each of the control variables is a correspondence between a certain attribute of a certain power system model and each of the data organization structures;

[0008] Based on the performance requirements of each of the power system models and each of the data organization structures, constructing the constraint conditions of the control variables and the objective function of the data organization structure selection;

[0009] A preset genetic algorithm is used to solve the problem according to each of the control variables, the constraint conditions, and the objective function to obtain a solved power system data organization structure matching scheme, wherein the power system data organization structure matching scheme is a correspondence between each power system model and each data organization structure in the power system;

[0010] According to the type of each power system model, each power system model and each data organization structure in the solved power system data organization structure matching scheme is merged to obtain a final power system data organization structure matching scheme.

[0011] The embodiment of the present application provides a data organization structure selection method for a power system model, which fully utilizes the strong global search capability of the genetic algorithm and is suitable for the optimization problems of complex functions such as nonlinear, non-convex, and multi-peak functions. Different power system data organization structure selection schemes are used as different individuals in the genetic algorithm. Through continuous iterative inheritance, individual optimization genetic solutions are performed. The genetic algorithm is cleverly combined with the data organization structure selection problem of each power system model to achieve automatic selection of the data organization structure of the power system model, avoiding the existing technology that mainly relies on the engineering experience of technicians to select the corresponding power system model data organization structure, and improving the selection efficiency of the data organization structure. On the other hand, the embodiment of the present application is based on the performance requirements of each of the power system models and each of the data organization structures, constructs the constraint conditions of the control variables and the objective function of the data organization structure selection, so that in the process of solving the problem using the genetic algorithm, the direction of the solution can be controlled by the constraint conditions and the objective function, ensuring that the final power system data organization structure selection scheme can meet the data retrieval performance requirements of the power grid control business in various operating scenarios.

[0012] Furthermore, according to each of the power system models and each of the data organization structures, a number of control variables are constructed, including:

[0013] generating all possible data organization structure matching schemes for each attribute in each power system model according to each power system model and each data organization structure, and representing each data organization structure matching scheme as a row vector;

[0014] The elements in the row vector are used to indicate whether a certain attribute of the power system model is matched with a certain data organization structure. The number of elements in the row vector is determined by the number of the data organization structures, and the value of each element is 0 or 1.

[0015] Compression and dimensionality reduction are performed on each of the row vectors to generate corresponding control variables.

[0016] Furthermore, compressing and reducing the dimension of each row vector to generate the corresponding control variables includes:

[0017] Compressing each row vector into corresponding numerical values ​​in a binary to decimal conversion manner;

[0018] The respective numerical values ​​are used as the respective control variables.

[0019] The embodiment of the present application provides a method for constructing several control variables. The control variables are constructed so that when a genetic algorithm is subsequently used, the population can be continuously updated and screened by the control variables to obtain the final optimization result. Therefore, the control variables need to contain information about the data organization structure matching scheme to facilitate the subsequent solution of the optimal power system data organization structure matching scheme. The embodiment of the present application first represents each of the data organization structure matching schemes as a row vector, and converts the correspondence between the attributes of the power system model and the data organization structure into a vector representation that can be understood by the computer, providing a data basis for the subsequent implementation of the genetic algorithm. In particular, the embodiment of the present invention fully considers that the genetic algorithm has a certain degree of randomness and that there may be accidental errors in the solution results. To address this problem, a solution is provided for merging the organizational structure matching results of similar model attributes to reduce accidental errors. Furthermore, the embodiment of the present application cleverly utilizes the conversion relationship between binary and decimal, converting the binary row vectors into specific decimal values, significantly compressing the data size without losing key information, and improving the convergence speed of the subsequent genetic algorithm.

[0020] In one possible implementation, constructing the constraint conditions for control variables and the objective function for data organization structure selection based on the performance requirements for each of the power system models and each of the data organization structures includes:

[0021] Constructing a data retrieval time constraint for a control variable based on the data retrieval transaction time of each attribute of each power system model and a preset retrieval time limit;

[0022] Constructing a value range constraint for a control variable according to preset restrictions on the number of options for the data organization structure for each attribute of each power system model;

[0023] According to the optimization goal of minimizing data redundancy, an objective function of data organization structure matching is constructed, wherein the data redundancy is the number of data organization structures in the power system data organization structure matching scheme.

[0024] The embodiment of the present application fully considers and balances the performance requirements of data retrieval transactions in the power grid control business and the low data redundancy requirements of data organization. By constructing data retrieval time constraints, it ensures that the power system data organization structure selection scheme after solution can meet the performance requirements of data retrieval transactions in the power grid control business; by constructing the value range constraints of the control variables, it ensures that the number of data organization structures selected for each power system model meets the preset range, avoiding the situation where a certain power system model has no corresponding data organization structure during the optimization process. At the same time, according to the optimization goal of minimizing data redundancy, the objective function of data organization structure selection is constructed, which reduces the redundancy of each model data while ensuring data retrieval performance, which is conducive to reducing the cost of data synchronization and management of power system models.

[0025] In one possible implementation, the method of using a preset genetic algorithm to solve the problem based on the control variables, the constraints, and the objective function to obtain a solved power system data organization structure selection scheme includes:

[0026] Determining the iterative initial value of the genetic algorithm according to each of the control variables, wherein the iterative initial value includes the number of initial populations and the power system data organization structure matching scheme corresponding to each initial individual in the initial population;

[0027] Determining the fitness function of the genetic algorithm according to the objective function;

[0028] Performing several iterative calculations based on the genetic algorithm according to the initial value of the iteration, the constraint conditions, and the fitness function until a preset number of iterations is reached to obtain a final population, performing crossover and mutation on the current population based on the control variable to obtain an updated population, and then screening the updated population according to the constraint conditions and the fitness function to determine the population to enter the next iterative calculation;

[0029] Calculating the fitness of each individual in the final population according to the fitness function, and then determining the optimal individual according to each fitness;

[0030] The data organization structure matching scheme corresponding to the optimal individual is extracted to obtain the solved power system data organization structure matching scheme.

[0031] Furthermore, determining the initial value of the iteration of the genetic algorithm according to each of the control variables includes:

[0032] Determining the population size according to the number of the control variables and a preset genetic algorithm population size range;

[0033] Dividing each of the initial individuals into first initial individuals or second initial individuals according to a preset ratio;

[0034] For the first initial individual, the control variable corresponding to each attribute of each power system model is set to the maximum value, forming a power system data organization structure matching scheme corresponding to the first initial individual;

[0035] For the second initial individual, after excluding the power system data organization structure matching scheme corresponding to the first initial individual, taking into account the convergence of the algorithm, the optimal data organization scheme of some power system models can be directly preset based on the matching experience of actual engineering and technical personnel, and a control variable is randomly selected for each attribute of the remaining power system models to form the power system data organization structure matching scheme corresponding to the second initial individual.

[0036] The embodiment of the present application provides a method for solving the problem through a genetic algorithm, combining the original genetic algorithm with the actual problem of the present application, using different power system data organization structure selection schemes as different individuals in the genetic algorithm, and randomly combining various control variables as individual chromosomes. During the genetic algorithm process, the current population is subjected to crossover and mutation operations according to the control variables, the chromosomes of each individual are changed, and a new population is generated. The new population is then screened according to the constraints and fitness function, and the chromosomes of the qualified individuals are inherited to the next generation, thereby realizing the combination of the genetic algorithm and the power system model. Furthermore, in the process of constructing the iterative initial value, the embodiment of the present application improves the genetic algorithm, generating some initial population individuals in a fully redundant data organization manner, rather than randomly generating all individuals, taking into account both the convergence speed of the algorithm and the global optimality of the results.

[0037] Furthermore, the fitness function of the genetic algorithm is determined according to the objective function, and the specific formula is:

[0038]

[0039] Where f(x) is the objective function, n represents the number of power system models, c i Represents the number of attributes of the ith power system model, num(X i ) represents the total number of data organization structures of all attributes of the i-th power system model, and F is the fitness function of the genetic algorithm.

[0040] In a second aspect, an embodiment of the present application provides a data organization structure selection system for a power system model, including an acquisition module, a control variable construction module, a constraint and objective function construction module, a genetic algorithm solution module, and a merging module;

[0041] The acquisition module is used to acquire several power system models and several optional data organization structures in the power system;

[0042] The control variable construction module is used to construct a plurality of control variables according to each of the power system models and each of the data organization structures, wherein each of the control variables is a correspondence between a certain attribute of a certain power system model and each of the data organization structures;

[0043] The constraint and objective function building module is used to build the constraint conditions of the control variables and the objective function of the data organization structure selection based on the performance requirements of each of the power system models and each of the data organization structures;

[0044] The genetic algorithm solving module is used to use a preset genetic algorithm to solve the control variables, the constraints and the objective function to obtain a solved power system data organization structure matching scheme, wherein the power system data organization structure matching scheme is a correspondence between each power system model and each data organization structure in the power system;

[0045] The merging module is used to merge the various power system models and the various data organization structures in the solved power system data organization structure matching scheme according to the type of each power system model to obtain a final power system data organization structure matching scheme.

[0046] Furthermore, the control variable construction module constructs a number of control variables according to each of the power system models and each of the data organization structures, including:

[0047] generating all possible data organization structure matching schemes for each attribute in each power system model according to each power system model and each data organization structure, and representing each data organization structure matching scheme as a row vector;

[0048] The elements in the row vector are used to indicate whether a certain attribute of the power system model is matched with a certain data organization structure. The number of elements in the row vector is determined by the number of the data organization structures, and the value of each element is 0 or 1.

[0049] Compression and dimensionality reduction are performed on each of the row vectors to generate corresponding control variables.

[0050] In one possible implementation, the genetic algorithm solving module uses a preset genetic algorithm to solve the control variables, the constraints, and the objective function to obtain a solved power system data organization structure selection scheme, including:

[0051] Determining the iterative initial value of the genetic algorithm according to each of the control variables, wherein the iterative initial value includes the number of initial populations and the power system data organization structure matching scheme corresponding to each initial individual in the initial population;

[0052] Determining the fitness function of the genetic algorithm according to the objective function;

[0053] Performing several iterative calculations based on the genetic algorithm according to the initial value of the iteration, the constraint conditions, and the fitness function until a preset number of iterations is reached to obtain a final population, performing crossover and mutation on the current population based on the control variable to obtain an updated population, and then screening the updated population according to the constraint conditions and the fitness function to determine the population to enter the next iterative calculation;

[0054] Calculating the fitness of each individual in the final population according to the fitness function, and then determining the optimal individual according to each fitness;

[0055] The data organization structure matching scheme corresponding to the optimal individual is extracted to obtain the solved power system data organization structure matching scheme. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A flowchart of a method for selecting a data organization structure for a power system model provided in an embodiment of the present application;

[0057] Figure 2 A schematic diagram of the correspondence between a power system model and a data organization structure in a method for selecting a data organization structure for a power system model provided in an embodiment of the present application;

[0058] Figure 3 A schematic diagram of an algorithm flow for solving a data organization structure of a power system model provided in an embodiment of the present application using a genetic algorithm;

[0059] Figure 4 A schematic diagram of changes in objective function values ​​during an iteration using a genetic algorithm in a method for selecting a data organization structure for a power system model provided in an embodiment of the present application;

[0060] Figure 5 A schematic diagram of data organization structure selection for a plant topology model in a method for selecting a data organization structure for a power system model provided in an embodiment of the present application;

[0061] Figure 6 A schematic diagram of data organization structure selection for a component topology model in a method for selecting a data organization structure for a power system model provided in an embodiment of the present application;

[0062] Figure 7 A schematic diagram of data organization structure selection for power electronics and control models in a method for selecting a data organization structure for a power system model provided in an embodiment of the present application;

[0063] Figure 8 A schematic diagram of data organization structure selection for a power generation model in a method for selecting a data organization structure for a power system model provided in an embodiment of the present application;

[0064] Figure 9 A schematic diagram of data organization structure selection for a load model in a method for selecting a data organization structure for a power system model provided in an embodiment of the present application;

[0065] Figure 10 A structural diagram of a data organization structure selection system for a power system model provided in an embodiment of the present application. DETAILED DESCRIPTION

[0066] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0067] It should be noted that the step numbers herein are for convenience of explanation of the specific embodiments and do not serve to define the order in which the steps are to be performed. In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature designated "first" or "second" may explicitly or implicitly include one or more of such features.

[0068] Example 1:

[0069] like Figure 1 As shown, embodiment 1 provides a method for selecting a data organization structure of a power system model, including steps S1-S5:

[0070] Step S1: obtaining a plurality of power system models and a plurality of optional data organization structures in the power system;

[0071] Step S2: constructing a plurality of control variables according to each of the power system models and each of the data organization structures, wherein each of the control variables is a correspondence between a certain attribute of a certain power system model and each of the data organization structures;

[0072] Step S3: constructing constraint conditions for control variables and an objective function for data organization structure selection based on performance requirements for each of the power system models and each of the data organization structures;

[0073] Step S4: using a preset genetic algorithm to solve the problem based on the control variables, the constraints, and the objective function to obtain a solved power system data organization structure matching scheme, wherein the power system data organization structure matching scheme is a correspondence between each power system model and each data organization structure in the power system;

[0074] Step S5: merging the power system models and the data organization structures in the solved power system data organization structure matching scheme according to the type of each power system model to obtain a final power system data organization structure matching scheme.

[0075] The embodiment of the present application provides a data organization structure selection method for a power system model, which fully utilizes the strong global search capability of the genetic algorithm and is suitable for the optimization problems of complex functions such as nonlinear, non-convex, and multi-peak functions. Different power system data organization structure selection schemes are used as different individuals in the genetic algorithm. Through continuous iterative inheritance, individual optimization genetic solutions are performed. The genetic algorithm is cleverly combined with the data organization structure selection problem of each power system model to achieve automatic selection of the data organization structure of the power system model, avoiding the existing technology that mainly relies on the engineering experience of technicians to select the corresponding power system model data organization structure, and improving the selection efficiency of the data organization structure. On the other hand, the embodiment of the present application is based on the performance requirements of each of the power system models and each of the data organization structures, constructs the constraint conditions of the control variables and the objective function of the data organization structure selection, so that in the process of solving the problem using the genetic algorithm, the direction of the solution can be controlled by the constraint conditions and the objective function, ensuring that the final power system data organization structure selection scheme can meet the data retrieval performance requirements of the power grid control business in various operating scenarios.

[0076] In a preferred embodiment, in step S1, as Figure 2 As shown in the figure, two power system models and one inter-model relationship are obtained. The two models include four and two attributes, respectively, and the inter-model relationship contains one attribute. In particular, the inter-model relationship can also be considered a special model with attributes and data organization structure, so it can be considered that there are three power system models. Figure 2In the example, three data organization structures, Structure A, Structure B, and Structure C, are provided. The total number of data organization structures is 3. Each of the above attributes can be used to select one or more data organization structures, thereby forming several power system data organization structure selection schemes. The present embodiment uses a genetic algorithm to automatically determine the optimal power system data organization structure selection scheme.

[0077] Furthermore, in step S2, a number of control variables are constructed according to each of the power system models and each of the data organization structures, including:

[0078] generating all possible data organization structure matching schemes for each attribute in each power system model according to each power system model and each data organization structure, and representing each data organization structure matching scheme as a row vector;

[0079] The elements in the row vector are used to indicate whether a certain attribute of the power system model is matched with a certain data organization structure. The number of elements in the row vector is determined by the number of the data organization structures, and the value of each element is 0 or 1.

[0080] Compression and dimensionality reduction are performed on each of the row vectors to generate corresponding control variables.

[0081] Furthermore, compressing and reducing the dimension of each row vector to generate the corresponding control variables includes:

[0082] Compressing each row vector into corresponding numerical values ​​in a binary to decimal conversion manner;

[0083] The respective numerical values ​​are used as the respective control variables.

[0084] The embodiment of the present application provides a method for constructing several control variables. The control variables are constructed so that when a genetic algorithm is subsequently used, the population can be continuously updated and screened by the control variables to obtain the final optimization result. Therefore, the control variables need to contain information about the data organization structure matching scheme to facilitate the subsequent solution of the optimal power system data organization structure matching scheme. The embodiment of the present application first represents each of the data organization structure matching schemes as a row vector, and converts the correspondence between the attributes of the power system model and the data organization structure into a vector representation that can be understood by the computer, providing a data basis for the subsequent implementation of the genetic algorithm. In particular, the embodiment of the present invention fully considers that the genetic algorithm has a certain degree of randomness and that there may be accidental errors in the solution results. To address this problem, a solution is provided for merging the organizational structure matching results of similar model attributes to reduce accidental errors. Furthermore, the embodiment of the present application cleverly utilizes the conversion relationship between binary and decimal, converting the binary row vectors into specific decimal values, significantly compressing the data size without losing key information, and improving the convergence speed of the subsequent genetic algorithm.

[0085] In a preferred embodiment, since the embodiment of the present application executes the genetic algorithm by a computer, it is necessary to abstract a specific power system data organization structure matching scheme into data that can be understood by the computer for representation. The embodiment of the present application represents a specific power system data organization structure matching scheme by establishing a matrix X, and the specific structure of X is as follows:

[0086]

[0087] Among them, x ij k Indicates the relationship between the j-th attribute of the i-th model and the k-th data organization scheme. A value of 1 indicates that the attribute adopts the k-th organization scheme, and a value of 0 indicates that it does not adopt it; n is the number of models; c i is the number of attributes in the i-th model; D is the number of data organization schemes. Therefore, each row vector in the X matrix represents a data organization structure matching scheme for a certain attribute in a power system model. Furthermore, a certain attribute in a power system model may have several possible data organization structure matching schemes. All of these matching schemes are generated and represented by different row vectors. Each row vector is then compressed and reduced in dimensionality to generate the corresponding control variables.

[0088] Furthermore, if all row vectors of the matrix X are compressed and integrated into a row vector in order from top to bottom Then we can get the chromosome code of a population individual in the genetic algorithm and prepare data for the subsequent genetic algorithm. Figure 2Taking the power system data organization structure selection scheme shown as an example, Figure 2 After the dimensionality reduction of the control variables in the , we can get the chromosome code Y = [x 11 ,x 12 ,x 13 ,x 14 ,x 21 ,x 22 ,x 31 ], where x 11 、x 12 、x 13 、x 14 They represent the decimal numbers after the data organization row vectors of the four attributes in model 1 are compressed, corresponding to x 21 、x 22 They represent the decimal numbers after the data organization row vectors of the two attributes in model 2 are compressed, and x 31 It represents the decimal number after the data organization row vector of one attribute in the relationship between models is compressed.

[0089] In one possible implementation, in step S3, based on the performance requirements of each of the power system models and each of the data organization structures, constructing the constraint conditions of the control variables and the objective function of the data organization structure selection includes:

[0090] Constructing a data retrieval time constraint for a control variable based on the data retrieval transaction time of each attribute of each power system model and a preset retrieval time limit;

[0091] Constructing a value range constraint for a control variable according to preset restrictions on the number of options for the data organization structure for each attribute of each power system model;

[0092] According to the optimization goal of minimizing data redundancy, an objective function of data organization structure matching is constructed, wherein the data redundancy is the number of data organization structures in the power system data organization structure matching scheme.

[0093] The embodiment of the present application fully considers and balances the performance requirements of data retrieval transactions in the power grid control business and the low data redundancy requirements of data organization. By constructing data retrieval time constraints, it ensures that the power system data organization structure selection scheme after solution can meet the performance requirements of data retrieval transactions in the power grid control business; by constructing the value range constraints of the control variables, it ensures that the number of data organization structures selected for each power system model meets the preset range, avoiding the situation where a certain power system model has no corresponding data organization structure during the optimization process. At the same time, according to the optimization goal of minimizing data redundancy, the objective function of data organization structure selection is constructed, which reduces the redundancy of each model data while ensuring data retrieval performance, which is conducive to reducing the cost of data synchronization and management of power system models.

[0094] In a preferred embodiment, the data retrieval time constraint, that is, the time consumption of each data retrieval transaction in each power grid control business scenario should not exceed its retrieval time limit, and the time consumption of each data retrieval transaction is approximately represented by the linear summation of one or more basic query requests, wherein the basic query requests include unique identification query, limited attribute range query, association relationship query and time series query;

[0095] Data value range constraint, that is, all power system model attributes should select at least one data organization scheme, and the control variable after dimensionality reduction is in the form of x ij , then x ij The value range of x is constrained to ij >0;

[0096] Furthermore, the penalty function is used to process the above data retrieval time constraint condition, and the penalty function max(0,(T l -Q l ) / Q l ), where T l Indicates the time taken for data retrieval transaction, Q l represents the corresponding retrieval time limit, then the penalty term of the data retrieval time constraint is: T l With Q l The larger value of the relative rate of change compared to 0.

[0097] Specifically, if the data retrieval transaction of this embodiment is DS1, and DS1 includes three basic query requests: one unique identifier query and two limited attribute range queries, as shown in Table 1, the time consumption of the four types of basic query requests is set respectively, thus T l =20+2*100, a total of 220ms, the corresponding data retrieval time constraint can be expressed as 220 l , the corresponding penalty function is max(0,(220-Q l ​) / Q l ); For the data value range constraint, the chromosome encodes all elements in Y, that is, x 11 、x 12 、x 13 、x 14 、x 21 、x 22 、x 31 Both must be greater than 0.

[0098] Table 1 Basic query requests and their time consumption

[0099] Serial number Basic query request type Time consumed / ms 1 Unique ID query time 10 2 Limiting the query time of attribute range 10 3 Time-consuming relationship query 10 4 Time series query time consumption 100

[0100] The objective function is defined as minimizing data redundancy, which is represented by the total number of data organization schemes for all attributes of the power system model. The greater the total number of data organization schemes, the greater the data redundancy of the selected scheme. Specifically:

[0101]

[0102] Among them, X i It is the submatrix of the matrix X for the i-th model data, and its number of columns is the same as X; the function num(X i ) Calculate the matrix X i The number of non-zero elements of represents the total number of data organization schemes for all attributes of the i-th model; n is the number of models; c i is the number of attributes of the i-th model. The fitness function of the genetic algorithm is defined as F, and its value is set to the inverse of the objective function, which can be obtained

[0103] Specifically, if Figure 2 As shown, in this embodiment, the total number n of relationships between models is 3, and the corresponding number of attributes c i There are 4, 2, and 1 items respectively. From this, we can conclude that the optimization goal of the current model data organization problem is:

[0104]

[0105] Among them, X i It is the submatrix of the matrix X for the i-th model data, and its number of columns is the same as X; the function num(X i ) Calculate the matrix X i The number of non-zero elements of represents the total number of data organization schemes for all attributes of the i-th model; accordingly, the fitness function of the genetic algorithm can be obtained as

[0106] In one possible implementation, in step S4, solving the problem using a preset genetic algorithm based on the control variables, the constraints, and the objective function to obtain a solved power system data organization structure selection scheme includes:

[0107] Determining the iterative initial value of the genetic algorithm according to each of the control variables, wherein the iterative initial value includes the number of initial populations and the power system data organization structure matching scheme corresponding to each initial individual in the initial population;

[0108] Determining the fitness function of the genetic algorithm according to the objective function;

[0109] Performing several iterative calculations based on the genetic algorithm according to the initial value of the iteration, the constraint conditions, and the fitness function until a preset number of iterations is reached to obtain a final population, performing crossover and mutation on the current population based on the control variable to obtain an updated population, and then screening the updated population according to the constraint conditions and the fitness function to determine the population to enter the next iterative calculation;

[0110] Calculating the fitness of each individual in the final population according to the fitness function, and then determining the optimal individual according to each fitness;

[0111] The data organization structure matching scheme corresponding to the optimal individual is extracted to obtain the solved power system data organization structure matching scheme.

[0112] Furthermore, determining the initial value of the iteration of the genetic algorithm according to each of the control variables includes:

[0113] Determining the population size according to the number of the control variables and a preset genetic algorithm population size range;

[0114] Dividing each of the initial individuals into first initial individuals or second initial individuals according to a preset ratio;

[0115] For the first initial individual, the control variable corresponding to each attribute of each power system model is set to the maximum value, forming a power system data organization structure matching scheme corresponding to the first initial individual;

[0116] For the second initial individual, after excluding the power system data organization structure matching scheme corresponding to the first initial individual, taking into account the convergence of the algorithm, the optimal data organization scheme of some power system models can be directly preset based on the matching experience of actual engineering and technical personnel, and a control variable is randomly selected for each attribute of the remaining power system models to form the power system data organization structure matching scheme corresponding to the second initial individual.

[0117] The embodiment of the present application provides a method for solving the problem through a genetic algorithm, combining the original genetic algorithm with the actual problem of the present application, using different power system data organization structure selection schemes as different individuals in the genetic algorithm, and randomly combining various control variables as individual chromosomes. During the genetic algorithm process, the current population is subjected to crossover and mutation operations according to the control variables, the chromosomes of each individual are changed, and a new population is generated. The new population is then screened according to the constraints and fitness function, and the chromosomes of the qualified individuals are inherited to the next generation, thereby realizing the combination of the genetic algorithm and the power system model. Furthermore, in the process of constructing the iterative initial value, the embodiment of the present application improves the genetic algorithm, generating some initial population individuals in a fully redundant data organization manner, rather than randomly generating all individuals, taking into account both the convergence speed of the algorithm and the global optimality of the results.

[0118] In a preferred embodiment, in step S4, the genetic algorithm population size is preset in the range of 10 to 250 to ensure a suitable algorithm convergence speed; except for the model in which the data organization scheme is directly preset based on engineering experience, the remaining models use a random number generation method to select some initial individuals with a probability of 25% to make their data organization schemes fully redundant, that is, the control variables corresponding to each attribute of each power system model are set to the maximum value; the remaining 75% of individuals are randomly selected from possible data organization schemes (excluding schemes with fully redundant data organization) with equal probability to take into account both the convergence speed of the algorithm and the global optimality of the results.

[0119] Specifically, if Figure 2 As shown, in this embodiment, if the data organization scheme of model 2 is set to be fully redundant, the two attributes in model 2 select the three data organization structures A, B, and C. At this time, the data organization row vectors corresponding to the two attributes are all [1,1,1], and the corresponding chromosome code Y in x 21 =7,x 22 =7; the genetic algorithm population size is set to 100 to meet the demand.

[0120] like Figure 3 As shown in the figure, a process of solving the problem by genetic algorithm is presented. Figure 2 Taking the power system model and data organization structure shown as an example, in steps S1, S2 and S3, several control variables, constraints and objective functions for Model 1 and Model 2 are constructed, and then Figure 3The algorithm flow in [1] is used to solve the problem. After iterative calculations, the individuals with the lowest data redundancy are screened out, resulting in a power system data organization structure matching scheme that meets the query time requirements of data retrieval transaction DS1. In this case, data redundancy between Models 1 and 2 is minimized, and model data synchronization and management costs are also low. Finally, the aforementioned models are classified according to their characteristics. Models of the same type adopt the same organization scheme, resulting in a merged power system data organization structure matching scheme.

[0121] In a preferred embodiment, in step S5, after the solved power system data organization structure matching scheme is obtained through the genetic algorithm, each power system model is classified according to the model characteristics, and the organization structure matching results of the same type of model attributes are merged according to the principle of majority vote to obtain the final power system data organization structure matching scheme.

[0122] In order to further verify the effectiveness of the data organization structure selection method for the power system model provided in the embodiment of the present application, the following experiment was designed in one of the embodiments of the present application, specifically:

[0123] Regarding the definition of control variables and dimensionality reduction for selecting the data organization structure for power system models, this embodiment defines control variables as a data organization structure selection scheme for power system model attributes. Specifically, this embodiment considers plant and inter-plant line models, static component models such as busbars and lines, and dynamic component models such as synchronous generators and renewable energy generation, totaling 19 model types and 25 inter-model relationships. Each model has three types of attributes: basic attributes, parameter attributes, and state attributes. In particular, inter-model relationships can also be considered as a special model with attributes and data organization structures. Therefore, model-to-model relationships should be considered in the control variables, and the value of n should be 44. Furthermore, the number of data organization scheme types D = 4, including relational table structures, graph structures, cross-section table structures, and time series structures, with their storage media being relational databases, in-memory graph databases, relational databases, and time series databases, respectively.

[0124] Obviously, the dimension of the control variable matrix X is 132×4, and it needs to be reduced in dimension. In this embodiment, the row vector of the matrix X is used to represent the data organization structure matching scheme of a certain model attribute. The row vector [x ij 1 ,x ij 2 ,…,x ij D ]Compress it into a variable x according to the binary to decimal conversion method ij According to the above-mentioned dimension compression method, all row vectors of the matrix X are sequentially integrated into a row vector from top to bottom, and the chromosome encoding of the genetic algorithm can be obtained as Y=[x11 ,x 12 ,x 13 ,x 21 ,x 22 ,x 23 ,…,x 132 1,x 132 2,x 132 3], where the elements represent the compressed decimal numbers of the data organization row vectors referring to a single attribute of a model or relation.

[0125] In terms of constraint definition and processing, this embodiment takes into account the data retrieval time constraints and data value range constraints in the data organization structure selection problem of the power system model. As shown in Table 2, this embodiment lists typical data retrieval function scenarios and their time limit requirements in the power grid control business. The time consumption of a data retrieval transaction is approximately represented by the linear summation of one or more basic query requests. According to the number of basic query requests contained in each data retrieval transaction in the list, the data retrieval transaction time consumption T1, T2, ..., T 11 .

[0126] Data retrieval time constraint, that is, the time consumption of each data retrieval transaction in each power grid control business scenario should not exceed its retrieval time limit, then the time consumption of all data retrieval transactions in the list is T l Should be less than the corresponding retrieval time limit Q l , that is, it is necessary to satisfy T1<10,T2<20,T3<20,T4<100,…,T 11 <100; further, the penalty function is used to process the above data retrieval time constraint condition, and the penalty function max(0,(T l -Q l ) / Q l ), where T l Indicates the time taken for data retrieval transaction, Q l Represents the corresponding retrieval time limit, and the corresponding penalty functions are max(0,(T 11 -10) / 10),max(0,(T 11 -20) / 20),…,max(0,(T 11 -100) / 100).

[0127] Data value range constraint, that is, all power system model attributes select at least one data organization scheme, and the control variable after dimensionality reduction is in the form of x ij , then it can be concluded that all elements in the chromosome code Y satisfy

[0128] Table 2 Data retrieval functions and time limits for typical power grid operation scenarios

[0129] Serial number Data retrieval content Time limit / ms 1 Retrieve a list of plants, stations, lines and relationships 10 2 Retrieve basic properties and parameter attributes of the plant 20 3 Retrieve basic and parameter attributes of inter-station lines 20 4 Retrieve power flow distribution of power plant network 120 5 Search the component list in the factory site 30 6 Retrieve the cross-sectional status of each component in the plant 500 7 Retrieve a list of lines or short circuits 20 8 Retrieve the detailed parameter properties of a component 20 9 Retrieve component status within a limited time range 200 10 Retrieve the plants and stations and inter-station lines affected by the fault 1000 11 Screening abnormal component models and their parameters 100

[0130] In terms of problem objective definition and genetic algorithm fitness function definition, the goal of the data organization structure selection problem for the power system model in this embodiment is to minimize data redundancy. Data redundancy is characterized by the total number of data organization schemes for all attributes of the power system model. Since the total number of models considered in this embodiment is 44, and each model or relationship has three attributes, the optimization goal of the current model data organization problem can be derived as follows:

[0131]

[0132] Among them, X i It is the submatrix of the matrix X for the i-th model data, and its number of columns is the same as X; the function num(X i ) Calculate the matrix X i The number of non-zero elements of represents the total number of data organization schemes for all attributes of the i-th model; the corresponding fitness function of the genetic algorithm is taken as That's it.

[0133] In terms of setting the initial value of the genetic algorithm iteration, it is necessary to set the data organization scheme of a part of the initial individuals to be fully redundant, that is, all data organization schemes are selected, and the data organization schemes of the remaining individuals are randomly generated to take into account the convergence speed of the algorithm and the global optimality of the results. In this embodiment, in addition to the model of the data organization scheme directly preset based on engineering experience, the remaining models use a random number generation method to select a part of the data organization structure with a probability of 25% to be fully redundant. That is, in this model, 4 data organization structures are selected for all 3 attributes, and the data organization row vectors corresponding to all attributes are [1,1,1,1], and the corresponding elements in the chromosome code Y are [1,1,1,1]. The remaining 75% of the models have two possible data organization schemes: 4 -1 (excluding the fully redundant data organization scheme), that is, the data organization row vector corresponding to a single attribute of the model may be [0,0,0,1], [0,0,1,0], [0,1,0,0], [1,0,0,0], etc. 15 kinds, and one data organization scheme is randomly selected with an equal probability of 1 / 15; in addition, the individual size of the genetic algorithm population selected in this embodiment is set to 200, the maximum genetic generation is 1000, and the crossover probability and mutation probability are both 0.8.

[0134] In terms of using genetic algorithms to solve mathematical models, in this embodiment, an initialized population is constructed through the above-mentioned method of setting the initial value of the iteration. During the iterative calculation process, the following three operations are repeatedly performed: determining whether the control vector corresponding to the individual meets all constraints, calculating the fitness of the individual that meets all constraints, and selecting whether to inherit to the next generation according to the probability corresponding to the numerical value of the individual fitness result. Figure 4 The following figure shows the changing trend of the objective function value during the iteration process. In the process of continuous selection, crossover and mutation, the data retrieval performance requirements of all typical scenarios are met, all other constraints are satisfied, and the optimization target value is gradually reduced. After the calculation is completed, the optimal objective function value is 5.814, and the following is obtained: Figures 5 to 9 The data organization structure options after solution are shown.

[0135] Since genetic algorithms have a certain degree of randomness, the solution results also have some random factors. For example, Figure 8 The basic attributes of the synchronous generator model and the PSS model are equipped with a table structure, while the other generator side models are not equipped here. In order to shield the accidental errors caused by the solution algorithm, in actual application, the same organization scheme is adopted for the same model. The preliminary organization scheme of the attributes of the same model is merged according to the principle of majority vote (if the votes are the same, one is randomly selected). For example, in the power generation side model, only the basic attributes of the synchronous generator and PSS models are equipped with a table structure, which accounts for a minority of votes. Therefore, the basic attributes of all models on the power generation side are not organized according to the table structure. In this embodiment, the models involved in the preliminary data organization scheme are classified according to the characteristics of the power system components, and the following five categories can be obtained: plant topology models (plant station, inter-station line), component topology models (busbar, line, transformer, shunt reactance and shunt capacitor), power electronics and control models (DC equipment, power electronic equipment and controller), power generation models (synchronous generator, excitation system, prime mover and speed regulator, PSS model, new energy) and load models (constant load, static load and dynamic load).

[0136] Table 3 shows the selected organizational structure for the power system data after manual merging in this embodiment. As can be seen from the table, the basic attributes of all models adopt a graph structure to facilitate grid topology search operations, and the parameter attributes of all models adopt a relational table structure to meet the parameter retrieval requirements of individual models. For parameter attributes, except for power electronics and control models, the parameter attributes of all other models adopt a graph structure. This redundant organization scheme allows direct access to model parameters after completing topology search. Because the number of power electronics model entities in actual power grids is smaller than that of other models, the performance of continuing to search in relational tables after completing topology search can also meet the requirements, so the redundant organization method of a graph structure is not designed. For state attributes, the state of plant topology models, component topology models, and load models is primarily used to query the spatial distribution of grid state at a specific time section, so a cross-section table structure is selected. However, with the exception of the state of plant topology models, the state attributes of the remaining models are required for data analysis operations such as time series data processing, parameter identification, and model verification analysis. Therefore, corresponding time series structures are selected to meet the high time series data retrieval requirements.

[0137] Table 3. Selection of power system data organization structure after merger

[0138]

[0139] In summary, the data organization structure selection method for the power system model based on the improved genetic algorithm provided in the embodiment of the present application fully considers and balances the performance requirements of data retrieval transactions in the power grid control business and the low data redundancy requirements of data organization, and realizes the automatic selection of the data organization structure of the power system model through modeling. During the modeling process, the embodiment of the present invention improves the genetic algorithm, including reducing the dimensionality of the control variables by compressing binary to decimal, introducing a penalty function for data retrieval time constraints, and generating some initial population individuals in a fully redundant manner for data organization. The above-mentioned improved method takes into account the convergence speed of the algorithm and the global optimality of the results. In particular, the embodiment of the present application fully considers that the genetic algorithm has a certain degree of randomness and the solution results may have accidental errors. To address this problem, a solution is provided for manually merging the organizational structure selection results of similar model attributes according to the principle of majority vote. After demonstration and analysis, the method described in the embodiment of the present invention can optimize the selection of data organization schemes, has applicability and effectiveness, and while meeting the performance requirements of various typical scenarios of the control business, achieves minimum data redundancy, which is conducive to reducing the cost of data synchronization and management of power system models.

[0140] Example 2:

[0141] like Figure 10As shown, embodiment 2 provides a data organization structure selection system for a power system model, including an acquisition module 10, a control variable construction module 20, a constraint and objective function construction module 30, a genetic algorithm solution module 40, and a merging module 50;

[0142] The acquisition module 10 is used to acquire several power system models and several optional data organization structures in the power system;

[0143] The control variable construction module 20 is used to construct a plurality of control variables according to each of the power system models and each of the data organization structures, wherein each of the control variables is a correspondence between a certain attribute of a certain power system model and each of the data organization structures;

[0144] The constraint and objective function building module 30 is used to build the constraint conditions of the control variables and the objective function of the data organization structure selection based on the performance requirements of each of the power system models and each of the data organization structures;

[0145] The genetic algorithm solving module 40 is used to solve the problem using a preset genetic algorithm according to each of the control variables, the constraints, and the objective function to obtain a solved power system data organization structure matching scheme, wherein the power system data organization structure matching scheme is a correspondence between each power system model and each data organization structure in the power system;

[0146] The merging module 50 is used to merge the power system models and the data organization structures in the solved power system data organization structure matching scheme according to the type of each power system model to obtain a final power system data organization structure matching scheme.

[0147] Furthermore, the control variable construction module 20 constructs several control variables according to each of the power system models and each of the data organization structures, including:

[0148] generating all possible data organization structure matching schemes for each attribute in each power system model according to each power system model and each data organization structure, and representing each data organization structure matching scheme as a row vector;

[0149] The elements in the row vector are used to indicate whether a certain attribute of the power system model is matched with a certain data organization structure. The number of elements in the row vector is determined by the number of the data organization structures, and the value of each element is 0 or 1.

[0150] Compression and dimensionality reduction are performed on each of the row vectors to generate corresponding control variables.

[0151] Furthermore, compressing and reducing the dimension of each row vector to generate the corresponding control variables includes:

[0152] Compressing each row vector into corresponding numerical values ​​in a binary to decimal conversion manner;

[0153] The respective numerical values ​​are used as the respective control variables.

[0154] In one possible implementation, the constraint and objective function construction module 30 constructs the constraint conditions of the control variables and the objective function of the data organization structure selection based on the performance requirements of each of the power system models and each of the data organization structures, including:

[0155] Constructing a data retrieval time constraint for a control variable based on the data retrieval transaction time of each attribute of each power system model and a preset retrieval time limit;

[0156] Constructing a value range constraint for a control variable according to preset restrictions on the number of options for the data organization structure for each attribute of each power system model;

[0157] According to the optimization goal of minimizing data redundancy, an objective function of data organization structure matching is constructed, wherein the data redundancy is the number of data organization structures in the power system data organization structure matching scheme.

[0158] In one possible implementation, the genetic algorithm solving module 40 uses a preset genetic algorithm to solve the control variables, the constraints, and the objective function to obtain a solved power system data organization structure selection scheme, including:

[0159] Determining the iterative initial value of the genetic algorithm according to each of the control variables, wherein the iterative initial value includes the number of initial populations and the power system data organization structure matching scheme corresponding to each initial individual in the initial population;

[0160] Determining the fitness function of the genetic algorithm according to the objective function;

[0161] Performing several iterative calculations based on the genetic algorithm according to the initial value of the iteration, the constraint conditions, and the fitness function until a preset number of iterations is reached to obtain a final population, performing crossover and mutation on the current population based on the control variable to obtain an updated population, and then screening the updated population according to the constraint conditions and the fitness function to determine the population to enter the next iterative calculation;

[0162] Calculating the fitness of each individual in the final population according to the fitness function, and then determining the optimal individual according to each fitness;

[0163] The data organization structure matching scheme corresponding to the optimal individual is extracted to obtain the solved power system data organization structure matching scheme.

[0164] Furthermore, determining the initial value of the iteration of the genetic algorithm according to each of the control variables includes:

[0165] Determining the population size according to the number of the control variables and a preset genetic algorithm population size range;

[0166] Dividing each of the initial individuals into first initial individuals or second initial individuals according to a preset ratio;

[0167] For the first initial individual, the control variable corresponding to each attribute of each power system model is set to the maximum value, forming a power system data organization structure matching scheme corresponding to the first initial individual;

[0168] For the second initial individual, after excluding the power system data organization structure matching scheme corresponding to the first initial individual, taking into account the convergence of the algorithm, the optimal data organization scheme of some power system models can be directly preset based on the matching experience of actual engineering and technical personnel, and a control variable is randomly selected for each attribute of the remaining power system models to form the power system data organization structure matching scheme corresponding to the second initial individual.

[0169] The embodiment of the present application provides a data organization structure selection system for a power system model, which fully utilizes the strong global search capability of the genetic algorithm and is suitable for complex function optimization problems such as nonlinear, non-convex, and multi-peak functions. Different power system data organization structure selection schemes are used as different individuals in the genetic algorithm. Through continuous iterative genetics, individual optimization genetic solutions are performed. The genetic algorithm is cleverly combined with the data organization structure selection problem of each power system model to achieve automatic selection of the data organization structure of the power system model, avoiding the existing technology that mainly relies on the engineering experience of technicians to select the corresponding power system model data organization structure, improving the selection efficiency of the data organization structure, and providing a modeling analysis and evaluation method for the power system model data organization scheme. On the other hand, the embodiment of the present application constructs the constraint conditions of the control variables and the objective function of the data organization structure selection based on the performance requirements of each power system model and each data organization structure, so that the direction of the solution can be controlled by the constraint conditions and the objective function during the solution process using the genetic algorithm, ensuring that the final power system data organization structure selection scheme can meet the data retrieval performance requirements of the power grid control business in various operation scenarios.

[0170] The more detailed working principle and process flow of this embodiment can be referred to, but not limited to, the relevant records of the first embodiment.

[0171] The specific embodiments described above further illustrate the objectives, technical solutions, and beneficial effects of this application. It should be understood that the above descriptions are merely specific embodiments of this application and are not intended to limit the scope of protection of this application. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of this application by those skilled in the art should be included within the scope of protection of this application.

Claims

1. A method for selecting a data organization structure for a power system model, characterized in that: include: Acquire several power system models and several optional data organization structures in the power system; Constructing a plurality of control variables according to each of the power system models and each of the data organization structures, wherein each of the control variables is a correspondence between a certain attribute of a certain power system model and each of the data organization structures; Based on the performance requirements of each of the power system models and each of the data organization structures, constructing the constraint conditions of the control variables and the objective function of the data organization structure selection; A preset genetic algorithm is used to solve the problem according to each of the control variables, the constraint conditions, and the objective function to obtain a solved power system data organization structure matching scheme, wherein the power system data organization structure matching scheme is a correspondence between each power system model and each data organization structure in the power system; According to the type of each power system model, each power system model and each data organization structure in the solved power system data organization structure matching scheme is merged to obtain a final power system data organization structure matching scheme.

2. The method for selecting a data organization structure of a power system model according to claim 1, wherein: According to each of the power system models and each of the data organization structures, a number of control variables are constructed, including: generating all possible data organization structure matching schemes for each attribute in each power system model according to each power system model and each data organization structure, and representing each data organization structure matching scheme as a row vector; The elements in the row vector are used to indicate whether a certain attribute of the power system model is matched with a certain data organization structure. The number of elements in the row vector is determined by the number of the data organization structures, and the value of each element is 0 or 1. Compression and dimensionality reduction are performed on each of the row vectors to generate corresponding control variables.

3. The method for selecting a data organization structure of a power system model according to claim 2, wherein: The compressing and dimensionality reduction of each row vector to generate corresponding control variables includes: Compressing each row vector into corresponding numerical values ​​in a binary to decimal conversion manner; The respective numerical values ​​are used as the respective control variables.

4. The method for selecting a data organization structure of a power system model according to claim 1, wherein: The constituting of the constraint conditions of the control variables and the objective function of the data organization structure selection based on the performance requirements of each of the power system models and each of the data organization structures includes: Constructing a data retrieval time constraint for a control variable based on the data retrieval transaction time of each attribute of each power system model and a preset retrieval time limit; Constructing a value range constraint for a control variable according to preset restrictions on the number of options for the data organization structure for each attribute of each power system model; According to the optimization goal of minimizing data redundancy, an objective function of data organization structure matching is constructed, wherein the data redundancy is the number of data organization structures in the power system data organization structure matching scheme.

5. The method for selecting a data organization structure of a power system model according to claim 1, wherein: The method of using a preset genetic algorithm to solve the control variables, the constraint conditions, and the objective function to obtain a solved power system data organization structure selection scheme includes: Determining the iterative initial value of the genetic algorithm according to each of the control variables, wherein the iterative initial value includes the number of initial populations and the power system data organization structure matching scheme corresponding to each initial individual in the initial population; Determining the fitness function of the genetic algorithm according to the objective function; Performing several iterative calculations based on the genetic algorithm according to the initial value of the iteration, the constraint conditions, and the fitness function until a preset number of iterations is reached to obtain a final population, performing crossover and mutation on the current population based on the control variable to obtain an updated population, and then screening the updated population according to the constraint conditions and the fitness function to determine the population to enter the next iterative calculation; Calculating the fitness of each individual in the final population according to the fitness function, and then determining the optimal individual according to each fitness; The data organization structure matching scheme corresponding to the optimal individual is extracted to obtain the solved power system data organization structure matching scheme.

6. The method for selecting a data organization structure of a power system model according to claim 5, wherein: Determining the initial value of the iteration of the genetic algorithm according to each of the control variables includes: Determining the population size according to the number of the control variables and a preset genetic algorithm population size range; Dividing each of the initial individuals into first initial individuals or second initial individuals according to a preset ratio; For the first initial individual, the control variable corresponding to each attribute of each power system model is set to the maximum value, forming a power system data organization structure matching scheme corresponding to the first initial individual; For the second initial individual, after excluding the power system data organization structure matching scheme corresponding to the first initial individual, a control variable is randomly selected for each attribute of each power system model to form the power system data organization structure matching scheme corresponding to the second initial individual.

7. The method for selecting a data organization structure of a power system model according to claim 5, wherein: The fitness function of the genetic algorithm is determined according to the objective function, and the specific formula is: Where f(x) is the objective function, n represents the number of power system models, c i Represents the number of attributes of the ith power system model, num(X i ) represents the total number of data organization structures of all attributes of the i-th power system model, and F is the fitness function of the genetic algorithm.

8. A data organization structure selection system for a power system model, characterized in that: It includes acquisition module, control variable construction module, constraint and objective function construction module, genetic algorithm solution module and merging module; The acquisition module is used to acquire several power system models and several optional data organization structures in the power system; The control variable construction module is used to construct a plurality of control variables according to each of the power system models and each of the data organization structures, wherein each of the control variables is a correspondence between a certain attribute of a certain power system model and each of the data organization structures; The constraint and objective function building module is used to build the constraint conditions of the control variables and the objective function of the data organization structure selection based on the performance requirements of each of the power system models and each of the data organization structures; The genetic algorithm solving module is used to use a preset genetic algorithm to solve the control variables, the constraints and the objective function to obtain a solved power system data organization structure matching scheme, wherein the power system data organization structure matching scheme is a correspondence between each power system model and each data organization structure in the power system; The merging module is used to merge the various power system models and the various data organization structures in the solved power system data organization structure matching scheme according to the type of each power system model to obtain a final power system data organization structure matching scheme.

9. A data organization structure selection system for a power system model according to claim 8, characterized in that: The control variable construction module constructs a number of control variables according to each of the power system models and each of the data organization structures, including: generating all possible data organization structure matching schemes for each attribute in each power system model according to each power system model and each data organization structure, and representing each data organization structure matching scheme as a row vector; The elements in the row vector are used to indicate whether a certain attribute of the power system model is matched with a certain data organization structure. The number of elements in the row vector is determined by the number of the data organization structures, and the value of each element is 0 or 1. Compression and dimensionality reduction are performed on each of the row vectors to generate corresponding control variables.

10. The data organization structure selection system for a power system model according to claim 8, characterized in that: The genetic algorithm solving module uses a preset genetic algorithm to solve the control variables, the constraint conditions and the objective function to obtain a solved power system data organization structure matching scheme, including: Determining the iterative initial value of the genetic algorithm according to each of the control variables, wherein the iterative initial value includes the number of initial populations and the power system data organization structure matching scheme corresponding to each initial individual in the initial population; Determining the fitness function of the genetic algorithm according to the objective function; Performing several iterative calculations based on the genetic algorithm according to the initial value of the iteration, the constraint conditions, and the fitness function until a preset number of iterations is reached to obtain a final population, performing crossover and mutation on the current population based on the control variable to obtain an updated population, and then screening the updated population according to the constraint conditions and the fitness function to determine the population to enter the next iterative calculation; Calculating the fitness of each individual in the final population according to the fitness function, and then determining the optimal individual according to each fitness; The data organization structure matching scheme corresponding to the optimal individual is extracted to obtain the solved power system data organization structure matching scheme.

Images