A distribution automation control system based on fault location

Abstract

The present invention relates to the adaptive control of a distribution network, and specifically to a distribution automation control system based on fault location. Fault signal acquisition terminals are distributively arranged at key nodes of the distribution network to collect fault signals and upload them to the fault location module. The fault location module determines the fault location by analyzing the fault signals. The adaptive control unit generates an optimal adaptive control strategy using an improved multi-objective starfish optimization algorithm according to the fault location and converts it into corresponding control instructions to be sent to the control instruction execution terminals. The control instruction execution terminals are distributively arranged at each node of the distribution network to execute the control instructions. The technical solution provided by the present invention can effectively overcome the defect that it is difficult to perform precise and timely adaptive control on the distribution network in the prior art.

Description

Technical Field

[0001] The present invention relates to the adaptive control of a distribution network, and particularly to a distribution automation control system based on fault location. Background Art

[0002] With the continuous development of the power system and the increasing growth of power demand, the security and reliability of the distribution network have become the focus of attention of power enterprises. However, in the complex and changeable distribution environment, the dynamic characteristics and disturbance factors of the distribution network are often difficult to accurately predict and model. Traditional distribution control methods, such as feedback control or optimal control based on fixed models, are difficult to adapt to this uncertainty and variability, resulting in unsatisfactory control effects and even potentially triggering new faults.

[0003] As a control method that can modify its own characteristics to adapt to the dynamic characteristics changes of the object and disturbances, adaptive control provides an effective way to solve the above problems. The research object of adaptive control is a system with a certain degree of uncertainty. It performs online identification of the system based on the input and output of the object. Through online identification, the adaptive control system can optimize the control strategy to ensure that the system operates in an optimal or sub-optimal state in a certain sense.

[0004] In the distribution automation control system, the application of adaptive control technology is particularly important. On the one hand, the complexity and uncertainty of the distribution network make it difficult for traditional control methods to achieve ideal control effects; on the other hand, adaptive control can adapt to the changes of the distribution network in real time and improve the control effect. However, there are still many deficiencies in the existing distribution control systems. For example, the adaptive ability is limited and it is difficult to meet the distribution control requirements under extreme conditions; the complexity and computational amount of the adaptive algorithm are large, affecting the real-time performance and stability of the system; and the integration of adaptive control with other control strategies is not tight enough, resulting in the need to improve the overall control effect, etc. Summary of the Invention

[0005] In view of the above-mentioned drawbacks of the prior art, the present invention provides a distribution automation control system based on fault location, which can effectively overcome the defect that it is difficult to perform accurate and timely adaptive control on the distribution network existing in the prior art.

[0006] To achieve the above object, the present invention is realized through the following technical solutions:

[0007] A distribution automation control system based on fault location, comprising:

[0008] Fault signal acquisition terminals, which are distributed at key nodes of the distribution network, collect fault signals and upload them to the fault location module;

[0009] A fault location module, which determines the fault location by analyzing the fault signals;

[0010] An adaptive control unit that uses an improved multi-objective starfish optimization algorithm to generate an optimal adaptive control strategy according to the fault location and converts it into corresponding control instructions to be sent to the control instruction execution terminal;

[0011] The control instruction execution terminal is distributed at each node of the distribution network and executes the control instructions;

[0012] Among them, the multi-gradient descent algorithm is used to provide an initial population for the improved multi-objective starfish optimization algorithm to provide an initial gradient direction and accelerate the convergence of the algorithm;

[0013] In the improved multi-objective starfish optimization algorithm, when simulating the predation behavior of starfish:

[0014] Dynamically adjust the step size according to the distance between the current solution and the optimal solution. When far from the optimal solution, use a larger step size to quickly approach. When close to the optimal solution, use a smaller step size for fine search;

[0015] Combine the gradient information of the objective function to guide the algorithm to search in the direction of gradient descent to accelerate the convergence of the algorithm;

[0016] Move towards the historical optimal solution direction with a certain probability to accelerate the convergence to the high-quality solution area;

[0017] Dynamically adjust the learning factor according to the number of iterations so that the algorithm focuses on global search in the initial stage and local development in the later stage.

[0018] Preferably, the fault signal acquisition terminal includes a fault indicator FI and a data acquisition terminal;

[0019] The fault indicator FI is distributed at key nodes of the distribution network, monitors voltage and current abnormalities in real time, and sends the traveling wave signal generated at the fault point to the data acquisition terminal;

[0020] The data acquisition terminal is distributed in the distribution network and uploads the collected traveling wave signals to the fault location module;

[0021] Among them, the key nodes of the distribution network include substations, branch boxes, distribution transformers, and cable joints.

[0022] Preferably, the fault location module determines the fault location by analyzing the fault signal, including:

[0023] Analyze the traveling wave signal generated at the fault point, calculate the fault location by measuring the time difference of the traveling wave reaching different fault indicators FI, and determine the fault location.

[0024] Preferably, the adaptive control unit generates an optimal adaptive control strategy according to the fault location by using an improved multi-objective starfish optimization algorithm, and converts it into corresponding control instructions to be sent to the control instruction execution terminal, including:

[0025] S1. Determine the objective function of distribution automation control;

[0026] S2. Determine the constraint conditions of distribution automation control;

[0027] S3. Combine the objective function and constraint conditions of distribution automation control to construct a distribution automation control model;

[0028] S4. Use the multi-gradient descent algorithm to perform multi-objective optimization and solution on the distribution automation control model, and use the generated initial solution set as the initial population in the improved multi-objective starfish optimization algorithm to provide the initial gradient direction and accelerate the algorithm convergence;

[0029] S5. Use the improved multi-objective starfish optimization algorithm to perform multi-objective optimization and solution on the distribution automation control model, and convert the generated optimal adaptive control strategy into corresponding control instructions to be sent to the control instruction execution terminal.

[0030] Preferably, S1. Determine the objective function of distribution automation control, including:

[0031] According to the power outage time function f 1( x ), the voltage deviation function f 2( x ), and the switch operation times function f 3( x ), determine the objective function of distribution automation control F :

[0032] ;

[0033] Wherein, x is the control variable, including the switch state, , , are the weight coefficients of the power outage time function f 1( x ), the voltage deviation function f 2( x ), and the switch operation times function f 3( x ) respectively;

[0034] S2. Determine the constraint conditions of distribution automation control, including:

[0035] Regard the distribution network topology constraint, equipment capacity limit and node voltage range as the constraint conditions of distribution automation control.

[0036] Preferably, in S4, a multi - gradient descent algorithm is used to perform multi - objective optimization on the distribution automatic control model, and the generated initial solution set is used as the initial population in the improved multi - objective starfish optimization algorithm to provide an initial gradient direction and accelerate the algorithm convergence, including:

[0037] S41. Calculate the gradients of the power outage time function f 1( x )), the voltage deviation function f 2( x ), and the switch operation times function f 3( x ) respectively, as follows: :

[0038] ;

[0039] Wherein, D is the dimension of the control variable, i is the function serial number, i = 1, 2, 3, , , are the gradients of the power outage time function f 1( x ), the voltage deviation function f 2( x ), and the switch operation times function f 3( x ) respectively;

[0040] S42. Construct the gradient matrix J :

[0041] ;

[0042] S43. Use the eigenvalue decomposition method to solve the following optimization problem to obtain the common gradient direction d * :

[0043] ;

[0044] Wherein, d is the direction vector, , represents taking the norm;

[0045] S44. Update the control variable d * along the common gradient direction to obtain the initial solution set x 0: x 0:

[0046] ;

[0047] Among them, is the first step size factor;

[0048] S45. Use the initial solution set x 0 as the initial population in the improved multi-objective starfish optimization algorithm to provide the initial gradient direction and accelerate the algorithm convergence.

[0049] Preferably, in S5, use the improved multi-objective starfish optimization algorithm to perform multi-objective optimization on the distribution automation control model, and convert the generated optimal adaptive control strategy into corresponding control instructions and send them to the control instruction execution terminal, including:

[0050] S51. In the search space, use the initial solution set as the initial population. Each starfish individual in the initial population represents a potential solution, and perform parameter initialization;

[0051] S52. In each iteration, optimize the solution quality by simulating the exploration behavior, predation behavior, and regeneration behavior of starfish;

[0052] S53. Use the non-dominated sorting algorithm to sort the starfish individuals in the population, calculate the crowding distance of each starfish individual, and select high-quality starfish individuals to enter the next generation population according to non-dominated sorting and crowding distance;

[0053] S54. Use the objective function of the distribution automation control F to evaluate each starfish individual in the population and calculate the corresponding objective function value;

[0054] S55. Record and update the optimal solution;

[0055] S56. Determine whether the iteration termination condition is satisfied. If the iteration termination condition is not satisfied, return to S52; otherwise, use the current optimal solution as the optimal adaptive control strategy;

[0056] S57. Convert the optimal adaptive control strategy into corresponding control instructions and send them to the control instruction execution terminal;

[0057] Among them, when simulating the predation behavior of starfish:

[0058] Dynamically adjust the step size according to the distance between the current solution and the optimal solution. When far from the optimal solution, use a larger step size to quickly approach; when close to the optimal solution, use a smaller step size for fine search;

[0059] Combine the gradient information of the objective function to guide the algorithm to search in the direction of gradient descent to accelerate the algorithm convergence;

[0060] Move towards the direction of the historical optimal solution with a certain probability to accelerate the convergence to the high-quality solution area;

[0061] Dynamically adjust the learning factor according to the number of iterations, so that the algorithm focuses on global search in the initial stage and local development in the later stage.

[0062] Preferably, in each iteration of S52, optimize the quality of the solution by simulating the exploration behavior, predation behavior, and regeneration behavior of starfish, including:

[0063] Simulate the starfish using its tentacles to sense the surrounding environment and move towards a better solution area:

[0064] ;

[0065] Wherein, P best is the position of the current optimal solution, , are respectively the positions of randomly selected starfish individuals X 1, X 2 at the k th iteration, is a random angle, , is the position of starfish individual j at the k +1th iteration;

[0066] Simulate the starfish moving towards the current optimal solution during predation, and at the same time consider the distance from other starfish individuals, so as to quickly converge to the current optimal solution and avoid local optima:

[0067] ;

[0068] Wherein, is the position of starfish individual j at the k th iteration, is a randomly selected starfish individual X 3 at the k th iteration, is the second step size factor of starfish individual j at the k th iteration, r 1 is a random number, ;

[0069] If a starfish individual has not moved its position for multiple consecutive generations, initialize its position by simulating the regeneration behavior of the starfish.

[0070] Preferably, when simulating the predation behavior of the starfish, dynamically adjust the step size according to the distance between the current solution and the optimal solution. When far from the optimal solution, use a larger step size to quickly approach, and when close to the optimal solution, use a smaller step size for fine search, including:

[0071] ;

[0072] Among them, is the distance between the starfish individual k at the j -th iteration and the current optimal solution, d max is the maximum distance between starfish individuals in the population, is the basic step size factor, is the step size adjustment speed factor;

[0073] When simulating the starfish predation behavior, the gradient information of the objective function is combined to guide the algorithm to search in the direction of gradient descent to accelerate the convergence of the algorithm, including:

[0074] ;

[0075] Among them, is the gradient of the objective function F at the position of the starfish individual j at the k -th iteration, is the learning rate;

[0076] When simulating the starfish predation behavior, it moves towards the historical optimal solution with a certain probability to accelerate the convergence to the high-quality solution region, including:

[0077] ;

[0078] Among them, P hbest is the position of the historical optimal solution, is the randomly selected starfish individual X 4 at the k -th iteration, is the guiding factor at the k -th iteration, r 2 is a random number, ;

[0079] When simulating the starfish predation behavior, the learning factor is dynamically adjusted according to the number of iterations, so that the algorithm focuses on global search in the initial stage and local development in the later stage, including:

[0080] For the second step size factor j of the starfish individual k at the -th iteration, it is dynamically adjusted according to the number of iterations using the following formula:

[0081] ;

[0082] Among them, k is the current iteration number, k max is the maximum iteration number, is the control factor, is the starfish individual j at the k +1-th iteration;

[0083] For the guiding factor k at the -th iteration, it is dynamically adjusted according to the iteration number by the following formula:

[0084] ;

[0085] Among them, is the guiding factor at the k +1-th iteration.

[0086] Preferably, in S53, the non-dominated sorting algorithm is used to sort the starfish individuals in the population, and the crowding distance of each starfish individual is calculated. High-quality starfish individuals are selected to enter the next-generation population according to non-dominated sorting and crowding distance, including:

[0087] S531. Use the fast non-dominated sorting algorithm in NSGA-II to sort the starfish individuals in the population;

[0088] S532. Calculate the crowding distance j of the starfish individual d j :

[0089] ;

[0090] Among them, F ( j +1), F ( j -1) are the objective function values of the starfish individuals j +1, j -1 adjacent to the starfish individual j F max min , F min are the maximum and minimum objective function values of the starfish individuals in the population, respectively;

[0091] S533. Select high-quality starfish individuals to enter the next-generation population according to non-dominated sorting and crowding distance.

[0092] Compared with the prior art, the distribution automation control system based on fault location provided by the present invention has the following beneficial effects:

[0093] 1) The multi - gradient descent algorithm is used to perform multi - objective optimization and solution on the distribution automation control model, and the generated initial solution set is used as the initial population in the improved multi - objective starfish optimization algorithm to provide the initial gradient direction for the improved multi - objective starfish optimization algorithm, accelerate its convergence, and effectively improve the response speed of the adaptive control unit.

[0094] 2) In the improved multi - objective starfish optimization algorithm, when simulating the starfish predation behavior, the step size is dynamically adjusted according to the distance between the current solution and the optimal solution. When far from the optimal solution, a larger step size is adopted to quickly approach; when close to the optimal solution, a smaller step size is adopted for fine - search. Combining with the gradient information of the objective function, the algorithm is guided to search in the direction of gradient descent to accelerate the algorithm convergence. Move in the direction of the historical optimal solution with a certain probability to accelerate the convergence to the high - quality solution area. Dynamically adjust the learning factor according to the number of iterations, so that the algorithm focuses on global search in the initial stage and local development in the later stage, enabling the adaptive control unit to effectively cope with the complexity and uncertainty of the distribution network, ensuring that the generated optimal adaptive control strategy can effectively adapt to the dynamic characteristic changes of the distribution network and disturbances, and at the same time can further improve the response speed of the adaptive control unit, so as to be able to perform precise and timely adaptive control on the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0096] Figure 1 is the system schematic diagram of the present invention;

[0097] Figure 2 is the operation flow schematic diagram of the adaptive control unit in the present invention;

[0098] Figure 3 is the flow schematic diagram of the adaptive control unit in the present invention using the improved multi - objective starfish optimization algorithm to perform multi - objective optimization and solution on the distribution automation control model. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0099] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0100] A distribution automation control system based on fault location, as Figure 1 shown, includes:

[0101] Fault signal acquisition terminals, which are distributed at key nodes of the distribution network, collect fault signals and upload them to the fault location module;

[0102] Fault location module, which determines the fault location by analyzing the fault signals;

[0103] Adaptive control unit, which uses an improved multi-objective starfish optimization algorithm to generate an optimal adaptive control strategy according to the fault location and converts it into corresponding control instructions to be sent to the control instruction execution terminal;

[0104] Control instruction execution terminals, which are distributed at each node of the distribution network and execute the control instructions;

[0105] Among them, the multi-gradient descent algorithm is used to provide an initial population for the improved multi-objective starfish optimization algorithm to provide an initial gradient direction and accelerate the convergence of the algorithm;

[0106] In the improved multi-objective starfish optimization algorithm, when simulating the predation behavior of starfish:

[0107] Dynamically adjust the step size according to the distance between the current solution and the optimal solution. When far from the optimal solution, use a larger step size to quickly approach; when close to the optimal solution, use a smaller step size for fine search;

[0108] Combine the gradient information of the objective function to guide the algorithm to search in the direction of gradient descent to accelerate the convergence of the algorithm;

[0109] Move towards the direction of the historical optimal solution with a certain probability to accelerate the convergence to the high-quality solution area;

[0110] Dynamically adjust the learning factor according to the number of iterations so that the algorithm focuses on global search in the initial stage and local development in the later stage.

[0111] The fault signal acquisition terminal includes a fault indicator FI and a data acquisition terminal;

[0112] Fault indicator FI, which is distributed at key nodes of the distribution network, monitors voltage and current anomalies in real time, and sends the traveling wave signals generated at the fault point to the data acquisition terminal;

[0113] Data acquisition terminal, which is distributed in the distribution network and uploads the collected traveling wave signals to the fault location module;

[0114] Among them, the key nodes of the distribution network include substations, branch boxes, distribution transformers, and cable joints.

[0115] The fault location module analyzes the fault signals to determine the fault location, including:

[0116] Analyze the traveling wave signals generated at the fault point, calculate the fault location by measuring the time difference of the traveling wave arriving at different fault indicators FI, so as to determine the fault location.

[0117] The adaptive control unit uses the improved multi-objective starfish optimization algorithm to generate the optimal adaptive control strategy according to the fault location, and converts it into corresponding control instructions to be sent to the control instruction execution terminal, as Figure 2 shown, including:

[0118] S1. Determine the objective function of distribution automation control;

[0119] S2. Determine the constraints of distribution automation control;

[0120] S3. Combine the objective function and constraints of distribution automation control to construct a distribution automation control model;

[0121] S4. Use the multi-gradient descent algorithm to perform multi-objective optimization on the distribution automation control model, and use the generated initial solution set as the initial population in the improved multi-objective starfish optimization algorithm to provide the initial gradient direction and accelerate the algorithm convergence;

[0122] S5. Use the improved multi-objective starfish optimization algorithm to perform multi-objective optimization on the distribution automation control model, and convert the generated optimal adaptive control strategy into corresponding control instructions to be sent to the control instruction execution terminal.

[0123] Specifically, S1. Determine the objective function of distribution automation control, including:

[0124] According to the power outage time function f 1( x ), voltage deviation function f 2( x ), and switch operation times function f 3( x ), determine the objective function of distribution automation control F :

[0125] ;

[0126] Among them, x is a control variable, including the switch state, , , are the weight coefficients of the power outage time function f 1( x ), the voltage deviation function f 2( x ), and the switch operation times function f 3( x ), respectively.

[0127] Specifically, S2, determine the distribution automation control constraints, including:

[0128] Regard the distribution network topology constraint, equipment capacity limit, and node voltage range as the distribution automation control constraints.

[0129] Specifically, S4, use the multi-gradient descent algorithm to perform multi-objective optimization on the distribution automation control model, and use the generated initial solution set as the initial population in the improved multi-objective starfish optimization algorithm to provide the initial gradient direction and accelerate the algorithm convergence, including:

[0130] S41, calculate the gradients f 1( x ), the voltage deviation function f 2( x ), and the switch operation times function f 3( x ) of the power outage time function :

[0131] ;

[0132] Among them, D is the dimension of the control variable, i is the function number, i = 1, 2, 3, , , are the gradients of the power outage time function f 1( x ), the voltage deviation function f 2( x ), and the switch operation times function f 3( x ), respectively;

[0133] S42, construct the gradient matrix J :

[0134] ;

[0135] S43. Solve the following optimization problem using the eigenvalue decomposition method to obtain the common gradient direction d * :

[0136] ;

[0137] wherein d is the direction vector, , denotes taking the norm;

[0138] S44. Update the control variable d * along the common gradient direction x to obtain the initial solution set x 0:

[0139] ;

[0140] wherein is the first step size factor;

[0141] S45. Use the initial solution set x 0 as the initial population in the improved multi-objective starfish optimization algorithm to provide the initial gradient direction and accelerate the convergence of the algorithm.

[0142] In the above technical solution, the multi-gradient descent algorithm is used to perform multi-objective optimization and solution on the distribution automatic control model, and the generated initial solution set is used as the initial population in the improved multi-objective starfish optimization algorithm to provide the initial gradient direction for the improved multi-objective starfish optimization algorithm and accelerate its convergence, effectively improving the response speed of the adaptive control unit.

[0143] Specifically, S5. Use the improved multi-objective starfish optimization algorithm to perform multi-objective optimization and solution on the distribution automatic control model, and convert the generated optimal adaptive control strategy into corresponding control instructions and send them to the control instruction execution terminal, such as Figure 3 shown, including:

[0144] S51. Use the initial solution set as the initial population in the search space. Each starfish individual in the initial population represents a potential solution, and parameter initialization is performed;

[0145] S52. In each iteration, optimize the quality of the solution by simulating the exploration behavior, predation behavior, and regeneration behavior of starfish;

[0146] S53. Use the non-dominated sorting algorithm to sort the starfish individuals in the population, calculate the crowding distance of each starfish individual, and select high-quality starfish individuals to enter the next-generation population according to non-dominated sorting and crowding distance;

[0147] S54. Objective Function for Distribution Automatic Control F Evaluate each starfish individual in the population and calculate the corresponding objective function value;

[0148] S55. Record and update the optimal solution;

[0149] S56. Determine whether the iteration termination condition is satisfied. If not, return to S52; otherwise, use the current optimal solution as the optimal adaptive control strategy;

[0150] S57. Convert the optimal adaptive control strategy into corresponding control instructions and send them to the control instruction execution terminal;

[0151] Among them, when simulating the starfish predation behavior:

[0152] Dynamically adjust the step size according to the distance between the current solution and the optimal solution. When far from the optimal solution, use a larger step size to quickly approach; when close to the optimal solution, use a smaller step size for fine search;

[0153] Combine the gradient information of the objective function to guide the algorithm to search in the direction of gradient descent to accelerate the convergence of the algorithm;

[0154] Move towards the direction of the historical optimal solution with a certain probability to accelerate the convergence to the high-quality solution area;

[0155] Dynamically adjust the learning factor according to the number of iterations so that the algorithm focuses on global search in the initial stage and local development in the later stage.

[0156] Specifically, in S52, in each iteration, optimize the quality of the solution by simulating the starfish exploration behavior, starfish predation behavior, and starfish regeneration behavior, including:

[0157] Simulate the starfish using its tentacles to sense the surrounding environment and move towards a better solution area:

[0158] ;

[0159] Among them, P best is the position of the current optimal solution, , are the randomly selected starfish individuals X 1, X 2 at the k th iteration, is a random angle, , is the position of the starfish individual j at the k +1th iteration;

[0160] By simulating the movement of starfish towards the current optimal solution during predation, while considering the distance between other starfish individuals, it converges rapidly towards the current optimal solution and avoids local optima:

[0161] ;

[0162] Among them, is the position of starfish individual j at the k th iteration, is the randomly selected starfish individual X 3 at the k th iteration, is the second step size factor of starfish individual j at the k th iteration, r 1 is a random number, ;

[0163] If a starfish individual has not moved its position for multiple consecutive generations, its position is initialized by simulating the starfish regeneration behavior.

[0164] 1) When simulating the starfish predation behavior, the step size is dynamically adjusted according to the distance between the current solution and the optimal solution. When far from the optimal solution, a larger step size is adopted to approach quickly, and when close to the optimal solution, a smaller step size is adopted for fine search, including:

[0165] ;

[0166] Among them, is the distance between starfish individual k at the j th iteration and the current optimal solution, d max is the maximum distance between starfish individuals in the population, is the basic step size factor, is the step size adjustment speed factor;

[0167] 2) When simulating the starfish predation behavior, the gradient information of the objective function is combined to guide the algorithm to search in the direction of gradient descent to accelerate the algorithm convergence, including:

[0168] ;

[0169] Among them, is the gradient of the objective function F at the position of starfish individual j at the k th iteration, is the learning rate;

[0170] 3) When simulating the predation behavior of starfish, move towards the direction of the historical optimal solution with a certain probability to accelerate the convergence towards the high-quality solution region, including:

[0171] ;

[0172] Among them, P hbest is the position of the historical optimal solution, is a randomly selected starfish individual X 4 is the position at the k th iteration, is the guiding factor at the k th iteration, r 2 is a random number, ;

[0173] 4) When simulating the predation behavior of starfish, dynamically adjust the learning factor according to the number of iterations, so that the algorithm focuses on global search in the initial stage and local development in the later stage, including:

[0174] For the starfish individual j at the k th iteration, the second step size factor , is dynamically adjusted according to the number of iterations using the following formula:

[0175] ;

[0176] Among them, k is the current number of iterations, k max is the maximum number of iterations, is the control factor, is the starfish individual j at the k +1 th iteration, the second step size factor;

[0177] For the guiding factor k at the th iteration, is dynamically adjusted according to the number of iterations using the following formula:

[0178] ;

[0179] Among them, is the guiding factor at the k +1 th iteration.

[0180] In the above technical solution, when improving the multi-objective starfish optimization algorithm and simulating the starfish predation behavior, the step size is dynamically adjusted according to the distance between the current solution and the optimal solution. When far from the optimal solution, a larger step size is adopted to quickly approach it. When approaching the optimal solution, a smaller step size is adopted for fine search. Combining with the gradient information of the objective function, the algorithm is guided to search in the direction of gradient descent to accelerate the convergence of the algorithm. Moving towards the direction of the historical optimal solution with a certain probability to accelerate the convergence towards the high-quality solution region. Dynamically adjusting the learning factor according to the number of iterations, so that the algorithm focuses on global search in the initial stage and local development in the later stage, enabling the adaptive control unit to effectively cope with the complexity and uncertainty of the distribution network, ensuring that the generated optimal adaptive control strategy can effectively adapt to the dynamic characteristic changes of the distribution network and disturbances, and at the same time being able to further improve the response speed of the adaptive control unit, thereby enabling precise and timely adaptive control of the distribution network.

[0181] Specifically, in S53, the non-dominated sorting algorithm is used to sort the starfish individuals in the population, and the crowding distance of each starfish individual is calculated. Selecting high-quality starfish individuals to enter the next-generation population according to non-dominated sorting and crowding distance, including:

[0182] S531: Using the fast non-dominated sorting algorithm in NSGA-II to sort the starfish individuals in the population;

[0183] S532: Calculating the crowding distance of the starfish individual j : d j :

[0184] ;

[0185] Among them, F ( j + 1), F ( j - 1) are respectively the objective function values of the starfish individuals j adjacent to the starfish individual j + 1, j - 1, F max , F min are respectively the maximum objective function value and the minimum objective function value of the starfish individuals in the population;

[0186] S533: Selecting high-quality starfish individuals to enter the next-generation population according to non-dominated sorting and crowding distance.

[0187] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A distribution automation control system based on fault location, characterized in that: Including: Fault signal acquisition terminals, which are distributed at key nodes of the distribution network, collect fault signals and upload them to the fault location module; Fault location module, which determines the fault location by analyzing the fault signals; Adaptive control unit, which uses an improved multi-objective starfish optimization algorithm to generate an optimal adaptive control strategy according to the fault location and converts it into corresponding control instructions to be sent to the control instruction execution terminal; Control instruction execution terminals, which are distributed at each node of the distribution network and execute the control instructions; Among them, the multi-gradient descent algorithm is used to provide an initial population for the improved multi-objective starfish optimization algorithm to provide an initial gradient direction and accelerate the algorithm convergence; In the improved multi-objective starfish optimization algorithm, when simulating the predation behavior of starfish: Dynamically adjust the step size according to the distance between the current solution and the optimal solution. When far from the optimal solution, use a large step size to quickly approach; when close to the optimal solution, use a small step size for fine search; Combine the gradient information of the objective function to guide the algorithm to search in the direction of gradient descent to accelerate the algorithm convergence; Move towards the direction of the historical optimal solution with a certain probability to accelerate the convergence to the high-quality solution area; Dynamically adjust the learning factor according to the number of iterations so that the algorithm focuses on global search in the initial stage and local development in the later stage.

2. The distribution automation control system based on fault location according to claim 1, wherein: The fault signal acquisition terminal includes a fault indicator FI and a data acquisition terminal; Fault indicator FI, which is distributed at key nodes of the distribution network, monitors voltage and current abnormalities in real time, and sends the traveling wave signal generated at the fault point to the data acquisition terminal; Data acquisition terminal, which is distributed in the distribution network and uploads the collected traveling wave signals to the fault location module; Among them, the key nodes of the distribution network include substations, branch boxes, distribution transformers, and cable joints.

3. The distribution automation control system based on fault location according to claim 2, characterized in that: The fault location module determines the fault location by analyzing the fault signals, including: Analyze the traveling wave signal generated at the fault point, calculate the fault location by measuring the time difference of the traveling wave reaching different fault indicators FI, and determine the fault location.

4. The distribution automation control system based on fault location according to claim 1, characterized in that: The adaptive control unit uses an improved multi-objective starfish optimization algorithm to generate an optimal adaptive control strategy according to the fault location and converts it into corresponding control instructions to be sent to the control instruction execution terminal, including: S1. Determine the objective function of distribution automation control; S2. Determine the constraint conditions of distribution automation control; S3. Combine the objective function and constraint conditions of distribution automation control to construct a distribution automation control model; S4. Use the multi-gradient descent algorithm to perform multi-objective optimization on the distribution automation control model, and use the generated initial solution set as the initial population in the improved multi-objective starfish optimization algorithm to provide an initial gradient direction and accelerate the algorithm convergence; S5. Use the improved multi-objective starfish optimization algorithm to perform multi-objective optimization on the distribution automation control model, and convert the generated optimal adaptive control strategy into corresponding control instructions to be sent to the control instruction execution terminal.

5. The distribution automation control system based on fault location according to claim 4, wherein: S1. Determine the objective function of distribution automation control, including: According to the power outage time function f1(x), voltage deviation function f2(x), and switch operation times function f3(x), determine the objective function F of distribution automation control: ； Among them, x is a control variable, including the switch state, , , are the weight coefficients of the power outage time function f1(x), the voltage deviation function f2(x), and the switch operation times function f3(x), respectively; S2. Determine the constraints for distribution automatic control, including: Regarding the distribution network topology constraints, equipment capacity limits, and node voltage ranges as the constraints for distribution automatic control.

6. The distribution automation control system based on fault location according to claim 4, wherein: S4. Use the multi-gradient descent algorithm to perform multi-objective optimization on the distribution automatic control model, and use the generated initial solution set as the initial population in the improved multi-objective starfish optimization algorithm to provide the initial gradient direction and accelerate the algorithm convergence, including: S41. Calculate the gradients of the power outage time function f1(x), the voltage deviation function f2(x), and the switch operation times function f3(x) respectively : ； where D is the dimension of the control variable, i is the function number, and i = 1, 2, 3, , , are the gradients of the power outage time function f1(x), the voltage deviation function f2(x), and the switch operation times function f3(x), respectively; S42. Construct the gradient matrix J: ； S43. Solve the following optimization problem using the eigenvalue decomposition method to obtain the common gradient direction d * : ； where d is the direction vector, , denotes taking the norm; S44. Along the common gradient direction d * Update the control variable x to obtain the initial solution set x0: ； Among them, is the first step factor; S45. Use the initial solution set x0 as the initial population in the improved multi-objective starfish optimization algorithm to provide the initial gradient direction and accelerate the algorithm convergence.

7. The distribution automation control system based on fault location according to claim 4, wherein: S5. Use the improved multi-objective starfish optimization algorithm to perform multi-objective optimization on the distribution automatic control model, and convert the generated optimal adaptive control strategy into corresponding control instructions and send them to the control instruction execution terminal, including: S51. In the search space, use the initial solution set as the initial population. Each starfish individual in the initial population represents a potential solution, and parameter initialization is performed; S52. In each iteration, optimize the solution quality by simulating the exploration behavior, predation behavior, and regeneration behavior of starfish; S53. Use the non-dominated sorting algorithm to sort the starfish individuals in the population, calculate the crowding distance of each starfish individual, and select high-quality starfish individuals to enter the next generation population according to non-dominated sorting and crowding distance; S54. Use the objective function F of distribution automatic control to evaluate each starfish individual in the population and calculate the corresponding objective function value; S55. Record and update the optimal solution; S56. Determine whether the iteration termination condition is satisfied. If the iteration termination condition is not satisfied, return to S52; otherwise, use the current optimal solution as the optimal adaptive control strategy; S57. Convert the optimal adaptive control strategy into corresponding control instructions and send them to the control instruction execution terminal; Among them, when simulating the predation behavior of starfish: Dynamically adjust the step size according to the distance between the current solution and the optimal solution. When far from the optimal solution, use a large step size to quickly approach; when close to the optimal solution, use a small step size for fine search; Combine the gradient information of the objective function to guide the algorithm to search in the direction of gradient descent to accelerate the algorithm convergence; Move in the direction of the historical optimal solution with a certain probability to accelerate the convergence to the high-quality solution area; Dynamically adjust the learning factor according to the number of iterations so that the algorithm focuses on global search in the initial stage and local development in the later stage.

8. The distribution automation control system based on fault location according to claim 7, characterized in that: S52. In each iteration, optimize the solution quality by simulating the exploration behavior, predation behavior, and regeneration behavior of starfish, including: Simulate the starfish using its tentacles to sense the surrounding environment and move towards a better solution area: ； Among them, P best is the position of the current optimal solution, , are the positions of randomly selected starfish individuals X1 and X2 at the k-th iteration, is a random angle, , is the position of starfish individual j at the (k + 1)-th iteration; Simulate the starfish moving towards the current optimal solution during predation, and at the same time consider the distance from other starfish individuals to quickly converge to the current optimal solution and avoid local optimality; ； Among them, is the position of starfish individual j at the k-th iteration, is the position of randomly selected starfish individual X3 at the k-th iteration, is the second step size factor of starfish individual j at the k-th iteration, r1 is a random number, ; If a starfish individual has not moved its position for multiple consecutive generations, then initialize the position of this starfish individual by simulating the regeneration behavior of starfish.

9. The distribution automation control system based on fault location according to claim 8, wherein: When simulating the predation behavior of starfish, the step size is dynamically adjusted according to the distance between the current solution and the optimal solution. When far from the optimal solution, a large step size is adopted to quickly approach it. When approaching the optimal solution, a small step size is adopted for fine search, including: ； Among them, is the distance between starfish individual j and the current optimal solution at the k-th iteration, d max is the maximum distance between starfish individuals in the population, is the basic step size factor, is the step size adjustment speed factor; When simulating the predation behavior of starfish, the gradient information of the objective function is combined to guide the algorithm to search in the direction of gradient descent to accelerate the convergence of the algorithm, including: ； Among them, is the gradient of the objective function F at the position of the starfish individual j at the k-th iteration, is the learning rate; When simulating the predation behavior of starfish, it moves towards the direction of the historical optimal solution with a certain probability to accelerate the convergence to the high-quality solution region, including: ； Among them, P hbest is the position of the historical optimal solution, is the position of the randomly selected starfish individual X4 at the k-th iteration, is the guiding factor at the k-th iteration, and r2 is a random number, ; When simulating the predation behavior of starfish, the learning factor is dynamically adjusted according to the number of iterations, so that the algorithm focuses on global search in the initial stage and local development in the later stage, including: For the second step size factor of starfish individual j at the k-th iteration , it is dynamically adjusted according to the number of iterations using the following formula: ； where k is the current iteration number, and k max is the maximum iteration number, is the control factor, is the second step length factor of starfish individual j at the (k + 1)-th iteration; For the guiding factor at the k-th iteration , it is dynamically adjusted according to the number of iterations using the following formula: ； Among them, is the guiding factor at the (k + 1)-th iteration.

10. The distribution automation control system based on fault location according to claim 7, characterized in that: S53. Use the non-dominated sorting algorithm to sort the starfish individuals in the population, calculate the crowding distance of each starfish individual, and select high-quality starfish individuals to enter the next-generation population according to non-dominated sorting and crowding distance, including: S531. Use the fast non-dominated sorting algorithm in NSGA-II to sort the starfish individuals in the population; S532. Calculate the crowding distance d of starfish individual j j : ； Among them, F(j + 1) and F(j - 1) are the objective function values of starfish individuals j + 1 and j - 1 adjacent to starfish individual j, respectively, and F max , F min are the maximum objective function value and the minimum objective function value of starfish individuals in the population, respectively; S533. Select high-quality starfish individuals to enter the next-generation population according to non-dominated sorting and crowding distance.

Images