Operation optimization method, device and system suitable for active power distribution network

Abstract

The invention discloses an operation optimization method, device and system suitable for an active power distribution network. The operation optimization method comprises the steps of obtaining a pre-constructed power distribution network optimization operation model; solving the power distribution network optimization operation model based on the obtained power flow data of the active power distribution network, and obtaining the photovoltaic optimal reactive power in the active power distribution network; wherein the power distribution network optimization operation model is constructed through the following method: taking the minimum network loss of the active power distribution network as an optimization target, and combining with a power distribution network flow analysis model to establish a target function; wherein the hyper-parameter of the power flow analysis model of the power distribution network is obtained by solving based on a differential evolution algorithm; constraint conditions are established by taking safe and stable operation of the active power distribution network as a target. The power supply reliability can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of distribution network analysis, and in particular relates to an operation optimization method, device and system applicable to an active distribution network. Background Art

[0002] With the widespread application of renewable energy, especially photovoltaic power generation, the operation of modern power systems faces new challenges. The optimization of active distribution networks has become increasingly important. Traditional distribution networks often lead to increased network losses and reduced power supply reliability under high penetration rates. At the same time, traditional power flow calculation methods cannot effectively deal with the problem of accurate power flow perception caused by the high proportion of new energy access. It is urgent to seek effective optimization methods to ensure the efficiency and safety of power supply.

[0003] In this context, data-driven power flow analysis is particularly necessary. By acquiring and processing power flow data in real time and combining it with advanced optimization algorithms, an operation optimization model for active distribution networks can be constructed. This approach can not only accurately minimize network losses, but also effectively monitor and ensure system stability. Data-driven decision-making will improve power supply reliability and provide strong support for the efficient integration of renewable energy, thereby promoting the sustainable development of the power system. Summary of the invention

[0004] In view of the above problems, the present invention proposes an operation optimization method, device and system applicable to an active power distribution network, which can improve power supply reliability.

[0005] In order to achieve the above technical objectives and the above technical effects, the present invention is implemented through the following technical solutions:

[0006] In a first aspect, the present invention provides an operation optimization method applicable to an active power distribution network, comprising:

[0007] Obtain pre-built distribution network optimization operation models;

[0008] Based on the acquired power flow data of the active distribution network, solving the distribution network optimization operation model to obtain the optimal reactive power of photovoltaics in the active distribution network;

[0009] The distribution network optimization operation model is constructed by the following method:

[0010] Taking the minimization of the network loss of the active distribution network as the optimization goal, the objective function is established in combination with the distribution network power flow analysis model; wherein the hyperparameters of the distribution network power flow analysis model are obtained based on the differential evolution algorithm;

[0011] With the goal of safe and stable operation of the active distribution network, constraints are established.

[0012] In combination with the first aspect, optionally, the distribution network power flow analysis model is a CatBoost machine learning model.

[0013] In combination with the first aspect, optionally, the method for obtaining the hyperparameters of the distribution network power flow analysis model includes:

[0014] Generate NP real parameter vectors of dimension D as the initial population, the initial population conforms to the uniform probability distribution, and the mathematical expression of the individuals in the initial population is:

[0015]

[0016] in,

[0017]

[0018] In the formula, is the i-th individual in the 0-generation population, is the jth variable in the i-th individual in the 0th generation population, are the upper and lower limits of the jth variable in the ith individual, respectively; rand(0,1) is a random value between 0 and 1, G is the population number, and D is the total number of hyperparameters of the distribution network power flow analysis model;

[0019] A mutation operation is performed on the population, and the formula used in the mutation operation is:

[0020]

[0021] Where V i G is the new individual obtained after the mutation operation, F is the scaling factor, ranging from 0 to 1, controlling Scaling; is the r1th individual of the Gth generation, is the r2th individual of generation G, is the r3th individual of generation G; r1 ranges from 1 to NP; r2 ranges from 1 to NP; r3 ranges from 1 to NP;

[0022] Execute the preset iterative process and determine whether the number of iterations reaches the set value. If it reaches the set value, end the iteration and output the optimal hyperparameter combination. The preset iterative process includes:

[0023] A crossover operation is performed on the new population obtained after the mutation operation. The formula used in the crossover operation is:

[0024]

[0025] In the formula, is the jth variable in the new i-th individual obtained after the crossover operation, and CR is the preset crossover probability; for The jth variable in the new i-th individual obtained after the mutation operation;

[0026] The new population obtained after the crossover operation is subjected to a selection operation, and the formula used in the selection operation is:

[0027]

[0028] in,

[0029]

[0030] In the formula, is the i-th individual in the G+1-th generation population, is the i-th individual after the G-th generation crossover operation, is the i-th individual after the G-th generation without crossover operation, f() is the hyperparameter optimization objective function, and the optimization goal of the hyperparameter optimization objective function is to minimize the root mean square error.

[0031] In combination with the first aspect, optionally, the objective function of the distribution network optimization operation model is expressed as:

[0032] min f=P net,loss (t)

[0033] in,

[0034] [P net,loss (t),V(t)]=f DE-CatBoost (P load (t),Q load (t),P PV (t),Q PV (t))

[0035] Where P net,loss (t) is the network loss of the active distribution network at time t; V(t) is the vector composed of the voltages of each node in the active distribution network at time t; P load (t) is the vector composed of the active load of each node in the active distribution network at time t; Q load (t) is the vector composed of reactive loads of each node in the active distribution network at time t; P PV (t) is the vector composed of the photovoltaic active power output of each node at time t; Q PV (t) is the vector composed of the photovoltaic reactive power output of each node at time t; f DE-CatBoost It is a power flow analysis model for distribution network.

[0036] In combination with the first aspect, optionally, the expression of the constraint condition of the distribution network optimization operation model includes:

[0037] V min ≤V k (t)≤V max

[0038]

[0039] Where V k (t) is the voltage of node k in the active distribution network at time t; V min 、V max are the lower and upper limits of node voltage in the active distribution network respectively; I jk (t) is the line current from node j to node k in the active distribution network at time t; is the maximum value of the line current from node j to node k in the active distribution network; is the maximum photovoltaic access capacity.

[0040] In combination with the first aspect, optionally, during the training process, the distribution network flow analysis model uses the active load of the active distribution network, the reactive load of the active distribution network, the photovoltaic active power output, and the photovoltaic reactive power output as input data; and uses the network loss of the active distribution network and the voltage of each node in the active distribution network as output data.

[0041] In combination with the first aspect, optionally, the distribution network optimization operation model is solved by a particle swarm algorithm.

[0042] In combination with the first aspect, optionally, the method for solving the distribution network optimization operation model includes:

[0043] Initialize the parameters of the particle swarm algorithm and set the maximum and minimum output values ​​of photovoltaic reactive power as the upper and lower limits of individuals in the particle swarm algorithm;

[0044] A set of photovoltaic reactive power output values ​​is randomly generated as individuals in the initial particle swarm;

[0045] The objective function of the distribution network optimization operation model is set as the fitness function of the particle swarm algorithm, and the fitness value of the initial particle swarm is calculated;

[0046] Compare the fitness value of each individual with the fitness value of the individual optimal solution and retain the optimal solution;

[0047] Update the speed and position of individuals according to the principle of particle swarm algorithm;

[0048] The optimal reactive power output of photovoltaics in the distribution network is obtained through continuous iteration.

[0049] In a second aspect, the present invention provides an operation optimization device applicable to an active power distribution network, comprising:

[0050] An acquisition module is used to obtain a pre-built distribution network optimization operation model;

[0051] An optimization module, for solving the distribution network optimization operation model based on the acquired power flow data of the active distribution network, and obtaining the optimal reactive power of photovoltaics in the active distribution network;

[0052] The distribution network optimization operation model is constructed by the following method:

[0053] Taking the minimization of the network loss of the active distribution network as the optimization goal, the objective function is established in combination with the distribution network power flow analysis model; wherein the hyperparameters of the distribution network power flow analysis model are obtained based on the differential evolution algorithm;

[0054] With the goal of safe and stable operation of the active distribution network, constraints are established.

[0055] In a third aspect, the present invention provides an operation optimization system applicable to an active power distribution network, comprising a storage medium and a processor;

[0056] The storage medium is used to store instructions;

[0057] The processor is configured to operate according to the instructions to execute the method according to any one of the first aspects.

[0058] Compared with the prior art, the present invention has the following beneficial effects:

[0059] The present invention utilizes the distribution network power flow analysis model to learn the mapping relationship between the distribution network state and the power flow distribution, without the need for iterative calculation of the Jacobian matrix, thus overcoming the problem of pathological power flow that the traditional power flow calculation method is prone to. The differential evolution algorithm is used to optimize the hyperparameters of the distribution network power flow analysis model to achieve optimal power flow performance. Finally, based on the distribution network optimization operation model, the optimal reactive power output based on distributed photovoltaics is achieved through network construction control, thereby enhancing the autonomy of the substation and improving the power supply reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative labor, among which:

[0061] Figure 1 A schematic flow chart of an operation optimization method applicable to an active power distribution network according to an embodiment of the present invention;

[0062] Figure 2 This is a schematic diagram of an application of a power flow analysis model for a distribution network according to an embodiment of the present invention;

[0063] Figure 3 A flow chart of hyperparameter optimization of a power distribution network flow analysis model according to an embodiment of the present invention;

[0064] Figure 4 A schematic diagram of a particle swarm optimization algorithm solution according to an embodiment of the present invention. DETAILED DESCRIPTION

[0065] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0066] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0067] Example 1

[0068] The present invention provides an operation optimization method applicable to an active power distribution network, such as Figure 1 As shown, the following steps are included:

[0069] (1) Obtain a pre-built distribution network optimization operation model;

[0070] (2) based on the acquired power flow data of the active distribution network, solving the distribution network optimization operation model to obtain the optimal reactive power of photovoltaics in the active distribution network;

[0071] The distribution network optimization operation model is constructed by the following method:

[0072] Taking the minimization of the network loss of the active distribution network as the optimization goal, the objective function is established in combination with the distribution network power flow analysis model; wherein the hyperparameters of the distribution network power flow analysis model are obtained based on the differential evolution algorithm;

[0073] With the goal of safe and stable operation of the active distribution network, constraints are established.

[0074] The CatBoost machine learning model is a Boosting ensemble learning model based on the improved GBDT algorithm proposed by scholar Liudmila. The CatBoost machine learning model also uses the additive model as a framework for multiple rounds of iterations to build multiple base learners, and uses a serial method to learn the residuals of training samples to gradually improve the accuracy of the base learners. The CatBoost machine learning model selects a completely symmetrical binary decision tree as the base learner of the model. A symmetrical tree is a decision tree in which all nodes in each layer select the same splitting condition, and the structure of the tree is always symmetrical. The symmetrical tree splits all nodes in each layer according to the feature with the greatest splitting benefit, builds layer by layer, and grows continuously until the number of layers of the tree reaches the depth threshold. During the calculation process, the traditional decision tree needs to traverse the entire tree to obtain the calculation result. However, since the completely symmetrical binary decision tree has the characteristics of the same-level splitting condition, the leaf node can be quickly retrieved by constructing a binary index, thereby greatly improving the calculation speed. Compared with the traditional machine learning algorithm that uses a single learner, the CatBoost machine learning model is to establish multiple weak learners, train the established weak learners with the training set data, and combine these weak learners with a certain combination strategy. Therefore, a strong learner with ideal effect can be obtained, thereby obtaining more accurate prediction results. Ensemble learning solves the overfitting problem when using a single model to a certain extent. Even if a weak learner makes a wrong prediction during the learning process, other weak learners can correct the error. Therefore, ensemble learning usually has better accuracy and generalization ability. To this end, in a specific implementation of an embodiment of the present invention, the distribution network flow analysis model is a CatBoost machine learning model.

[0075] The CatBoost machine learning model has the characteristics of low complexity, fast running speed and high accuracy, but it contains a large number of hyperparameters that affect the performance of the algorithm. The selection and setting of hyperparameters will directly affect the structure and performance of the optimization model. The trial and error configuration of the hyperparameters of the CatBoost machine learning model based solely on engineering experience is obviously random and lacks stability. The grid search and random search hyperparameter optimization methods consume a lot of computing costs for each function evaluation, and the optimization efficiency is difficult to meet the large number of hyperparameter configuration requirements in the CatBoost machine learning model. To this end, in a specific implementation of the embodiment of the present invention, Figure 2 As shown, the method for obtaining the hyperparameters of the distribution network power flow analysis model includes:

[0076] Generate NP real parameter vectors of dimension D as the initial population, the initial population conforms to the uniform probability distribution, and the mathematical expression of the individuals in the initial population is:

[0077]

[0078] in,

[0079]

[0080] In the formula, is the i-th individual in the 0th generation population, x i,j is the jth variable in the ith individual, are the upper and lower limits of the jth variable in the ith individual, respectively; rand(0,1) is a random value between 0 and 1, G is the population number, and D is the total number of hyperparameters of the distribution network power flow analysis model;

[0081] Execute the preset iterative process and determine whether the number of iterations reaches the set value. If it reaches the set value, end the iteration and output the optimal hyperparameter combination. The preset iterative process includes:

[0082] A mutation operation is performed on the population, and the formula used in the mutation operation is:

[0083]

[0084] Where V i G is the new individual obtained after the mutation operation, F is the scaling factor, ranging from 0 to 1, controlling Scaling; is the r1th individual of the Gth generation, is the r2th individual of generation G, is the r3th individual of generation G; r1 ranges from 1 to NP; r2 ranges from 1 to NP; r3 ranges from 1 to NP;

[0085] A crossover operation is performed on the new population obtained after the mutation operation. The formula used in the crossover operation is:

[0086]

[0087] In the formula, is the jth variable in the new i-th individual obtained after the crossover operation, and CR is the preset crossover probability; for The jth variable in the new i-th individual obtained after the mutation operation;

[0088] The new population obtained after the crossover operation is subjected to a selection operation, and the formula used in the selection operation is:

[0089]

[0090] in,

[0091]

[0092] In the formula, is the i-th individual in the G+1-th generation population, is the i-th individual after the G-th generation crossover operation, is the i-th individual after the G-th generation without crossover operation, f() is the hyperparameter optimization objective function, and the optimization goal of the hyperparameter optimization objective function is to minimize the root mean square error.

[0093] The present invention utilizes the distribution network power flow analysis model to learn the mapping relationship between the distribution network state and the power flow distribution, without the need for iterative calculation of the Jacobian matrix, thus overcoming the problem of pathological power flow that the traditional power flow calculation method is prone to. The differential evolution algorithm is used to optimize the hyperparameters of the distribution network power flow analysis model to achieve optimal power flow performance. Finally, based on the distribution network optimization operation model, the optimal reactive power output based on distributed photovoltaics is achieved through network construction control, thereby enhancing the autonomy of the substation and improving the power supply reliability.

[0094] In a specific implementation of the embodiment of the present invention, the expression of the objective function of the distribution network optimization operation model is:

[0095] min f=P net,loss (t)

[0096] in,

[0097] [P net,loss (t),V(t)]=f DE-CatBoost (P load (t),Q load (t),P PV (t),Q PV (t))

[0098] Where P net,loss (t) is the network loss of the active distribution network at time t; V(t) is the vector composed of the voltages of each node in the active distribution network at time t; P load (t) is the vector composed of the active load of each node in the active distribution network at time t; Q load (t) is the vector composed of reactive loads of each node in the active distribution network at time t; P PV (t) is the vector composed of the photovoltaic active power output of each node at time t; Q PV (t) is the vector composed of the photovoltaic reactive power output of each node at time t; f DE-CatBoost It is a power flow analysis model for distribution network.

[0099] The expression of the constraint condition of the distribution network optimization operation model includes:

[0100] V min ≤V k (t)≤V max

[0101]

[0102] Where V k (t) is the voltage of node k in the active distribution network at time t; V min 、V max are the lower and upper limits of node voltage in the active distribution network respectively; I jk (t) is the line current from node j to node k in the active distribution network at time t; is the maximum value of the line current from node j to node k in the active distribution network; is the maximum photovoltaic access capacity.

[0103] In a specific implementation of the embodiment of the present invention, during the training process, the distribution network power flow analysis model uses the active load of the active distribution network, the reactive load of the active distribution network, the photovoltaic active power output, and the photovoltaic reactive power output as input data; and uses the network loss of the active distribution network and the voltage of each node in the active distribution network as output data. For details, see Figure 3 .

[0104] In a specific implementation of the embodiment of the present invention, the distribution network optimization operation model is solved by a particle swarm algorithm. Figure 4 As shown, specifically, the method for solving the distribution network optimization operation model includes:

[0105] Initialize the parameters of the particle swarm algorithm and set the maximum and minimum output values ​​of photovoltaic reactive power as the upper and lower limits of particles in the particle swarm algorithm;

[0106] A set of photovoltaic reactive power output values ​​is randomly generated as the initial particle swarm;

[0107] The objective function of the distribution network optimization operation model is set as the fitness function of the particle swarm algorithm, and the fitness value of the initial particle swarm is calculated;

[0108] Compare the fitness value of each individual with the fitness value of the individual optimal solution and retain the optimal solution;

[0109] Update the speed and position of individuals according to the principle of particle swarm algorithm;

[0110] The optimal output of photovoltaic power in the distribution network is obtained through continuous iteration.

[0111] The formula used to update the particle's velocity and position is:

[0112]

[0113] In the formula, v h is the velocity vector of the hth particle, ω is the inertia weight, c1 and c2 are acceleration factors, r1 and r2 are random numbers uniformly distributed in the interval [0,1], and pbest h is the optimal position of the hth individual, x h is the position vector of the hth individual, gbest h is the optimal position of the h-th individual in the group.

[0114] In the embodiment of the present invention, a particle swarm algorithm is used to solve the distribution network optimization operation model, which can realize the optimization and improvement of the network analysis and operation status of the new flexible distribution network.

[0115] Example 2

[0116] Based on the same inventive concept as that of Embodiment 1, an operation optimization device applicable to an active power distribution network is provided in an embodiment of the present invention, comprising:

[0117] An acquisition module is used to obtain a pre-built distribution network optimization operation model;

[0118] An optimization module, for solving the distribution network optimization operation model based on the acquired power flow data of the active distribution network, and obtaining the optimal reactive power of photovoltaics in the active distribution network;

[0119] The distribution network optimization operation model is constructed by the following method:

[0120] Taking the minimization of the network loss of the active distribution network as the optimization goal, the objective function is established in combination with the distribution network power flow analysis model; wherein the hyperparameters of the distribution network power flow analysis model are obtained based on the differential evolution algorithm;

[0121] With the goal of safe and stable operation of the active distribution network, constraints are established.

[0122] The rest are the same as in Example 1.

[0123] Example 3

[0124] An embodiment of the present invention provides an operation optimization system applicable to an active power distribution network, including a storage medium and a processor;

[0125] The storage medium is used to store instructions;

[0126] The processor is configured to operate according to the instructions to execute the method described in embodiment 1.

[0127] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0128] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0129] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0130] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0131] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which all fall within the protection of the present invention.

[0132] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. An operation optimization method applicable to an active distribution network, characterized in that: include: Obtain pre-built distribution network optimization operation models; Based on the acquired power flow data of the active distribution network, solving the distribution network optimization operation model to obtain the optimal reactive power of photovoltaics in the active distribution network; The distribution network optimization operation model is constructed by the following method: Taking the minimization of the network loss of the active distribution network as the optimization goal, the objective function is established in combination with the distribution network power flow analysis model; wherein the hyperparameters of the distribution network power flow analysis model are obtained based on the differential evolution algorithm; With the goal of safe and stable operation of the active distribution network, constraints are established.

2. The operation optimization method applicable to an active power distribution network according to claim 1, characterized in that: The distribution network power flow analysis model is a CatBoost machine learning model.

3. An operation optimization method applicable to an active power distribution network according to claim 1 or 2, characterized in that: The method for obtaining the hyperparameters of the power distribution network flow analysis model includes: Generate NP real parameter vectors of dimension D as the initial population, the initial population conforms to the uniform probability distribution, and the mathematical expression of the individuals in the initial population is: in, In the formula, is the i-th individual in the 0-generation population, is the jth variable in the i-th individual in the 0-generation population, are the upper and lower limits of the jth variable in the ith individual, respectively; rand(0,1) is a random value between 0 and 1, G is the population algebra, and D is the total number of hyperparameters of the distribution network power flow analysis model; A mutation operation is performed on the population, and the formula used in the mutation operation is: In the formula, is the new individual obtained after the mutation operation, F is the scaling factor, ranging from 0 to 1, controlling Scaling; is the r1th individual of the Gth generation, is the r2th individual of generation G, is the r3th individual of generation G; r1 ranges from 1 to NP; r2 ranges from 1 to NP; r3 ranges from 1 to NP; Execute the preset iterative process and determine whether the number of iterations reaches the set value. If it reaches the set value, end the iteration and output the optimal hyperparameter combination. The preset iterative process includes: A crossover operation is performed on the new population obtained after the mutation operation. The formula used in the crossover operation is: In the formula, is the jth variable in the new i-th individual obtained after the crossover operation, and CR is the preset crossover probability; for The jth variable in the new i-th individual obtained after the mutation operation; The new population obtained after the crossover operation is subjected to a selection operation, and the formula used in the selection operation is: in, In the formula, is the i-th individual in the G+1-th generation population, is the i-th individual after the G-th generation crossover operation, is the i-th individual after the G-th generation without crossover operation, f() is the hyperparameter optimization objective function, and the optimization goal of the hyperparameter optimization objective function is to minimize the root mean square error.

4. The operation optimization method applicable to an active power distribution network according to claim 1, characterized in that: The expression of the objective function of the distribution network optimization operation model is: min f=P net,loss (t) in, [P net,loss (t),V(t)]=f DE-CatBoost (P load (t),Q load (t),P PV (t),Q PV (t)) Where P net,loss (t) is the network loss of the active distribution network at time t; V(t) is the vector composed of the voltages of each node in the active distribution network at time t; P load (t) is the vector composed of the active load of each node in the active distribution network at time t; Q load (t) is the vector composed of reactive loads of each node in the active distribution network at time t; P PV (t) is the vector composed of the photovoltaic active power output of each node at time t; Q PV (t) is the vector composed of the photovoltaic reactive power output of each node at time t; f DE-CatBoost It is a power flow analysis model for distribution network.

5. The operation optimization method applicable to an active power distribution network according to claim 4, characterized in that: The expression of the constraint condition of the distribution network optimization operation model includes: V min ≤V k (t)≤V max Where V k (t) is the voltage of node k in the active distribution network at time t; V min 、V max are the lower and upper limits of node voltage in the active distribution network respectively; I jk (t) is the line current from node j to node k in the active distribution network at time t; is the maximum value of the line current from node j to node k in the active distribution network; is the maximum photovoltaic access capacity.

6. The operation optimization method applicable to an active power distribution network according to claim 4, characterized in that: During the training process of the distribution network flow analysis model, the active load of the active distribution network, the reactive load of the active distribution network, the photovoltaic active power output, and the photovoltaic reactive power output are used as input data; and the network loss of the active distribution network and the voltage of each node in the active distribution network are used as output data.

7. The operation optimization method applicable to an active power distribution network according to claim 4, characterized in that: The distribution network optimization operation model is solved by particle swarm algorithm.

8. The operation optimization method applicable to an active power distribution network according to claim 7, characterized in that: The method for solving the distribution network optimization operation model includes: Initialize the parameters of the particle swarm algorithm and set the maximum and minimum output values ​​of photovoltaic reactive power as the upper and lower limits of individuals in the particle swarm algorithm; A set of photovoltaic reactive power output values ​​is randomly generated as individuals in the initial particle swarm; The objective function of the distribution network optimization operation model is set as the fitness function of the particle swarm algorithm, and the fitness value of the initial particle swarm is calculated; Compare the fitness value of each individual with the fitness value of the individual optimal solution and retain the optimal solution; Update the speed and position of individuals according to the principle of particle swarm algorithm; The optimal reactive power output of photovoltaics in the distribution network is obtained through continuous iteration.

9. An operation optimization device suitable for an active distribution network, characterized in that: include: An acquisition module is used to obtain a pre-built distribution network optimization operation model; An optimization module, for solving the distribution network optimization operation model based on the acquired power flow data of the active distribution network, and obtaining the optimal reactive power of photovoltaics in the active distribution network; The distribution network optimization operation model is constructed by the following method: Taking the minimization of the network loss of the active distribution network as the optimization goal, the objective function is established in combination with the distribution network power flow analysis model; wherein the hyperparameters of the distribution network power flow analysis model are obtained based on the differential evolution algorithm; With the goal of safe and stable operation of the active distribution network, constraints are established.

10. An operation optimization system suitable for an active distribution network, characterized in that: including storage media and processors; The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the method according to any one of claims 1-8.

Images