A method and system for optimizing the configuration of distribution network hierarchical protection points

Abstract

A method and system for optimizing the configuration of hierarchical protection point selection for a distribution network includes setting multiple objective functions and multiple constraints for protection point selection optimization, wherein the protection point selection optimization is to optimize the number and location of protection points; establishing a constraint violation degree calculation formula based on all constraints; obtaining distribution network parameters, and iteratively optimizing the multiple objective functions based on these parameters using a particle swarm algorithm that introduces a three-dimensional vector rotation matrix and a dynamic perturbation vector. During the optimization process, particles are screened based on the constraint violation degree, and particle priorities are calculated based on the constraint violation degree and congestion distance; a set number of protection points are randomly selected from the multiple protection points optimized, and the final protection points are obtained through clustering and screening. The present invention realizes three-dimensional exploration of the multidimensional target space of distribution network protection point selection, significantly improving the distribution uniformity and global search capability of the solution set.

Description

Technical Field

[0001] The present invention belongs to the technical field of distribution network protection, and more specifically, relates to a method and system for optimizing configuration of hierarchical protection points in a distribution network. Background Art

[0002] As a critical link in the power supply of power systems, the rationality and reliability of the distribution network's protection configuration directly determine the efficiency of fault isolation and the quality of power supply to users. With the continuous growth of load density, the complexity of grid topology, and the increasing demand for high power supply reliability, traditional protection point selection strategies based on manual experience are no longer able to meet the comprehensive protection requirements of modern distribution systems in terms of speed, selectivity, and economy.

[0003] Hierarchical protection refers to the classification of protection devices into different types based on line structure and load distribution, usually including trunk line protection and branch line protection, and coordinated configuration according to a certain action sequence and protection range. As a hierarchical protection configuration strategy, hierarchical protection achieves hierarchical fault current removal and action coordination by setting different levels of protection on the trunk line and branch line, thereby improving the selectivity and speed of fault handling. However, in current actual projects, hierarchical protection configuration generally has problems such as the lack of quantitative standards for protection location and quantity, insufficient protection action speed, and lack of overall economic constraints. A unified set of standards and systematic optimization methods has not yet been formed. In addition, actual projects rely heavily on manual experience and set up protection equipment nearby. There are problems such as redundancy of trunk line protection levels and lack of branch line protection, which exposes the limitations of traditional protection site selection methods and affects the coordinated optimization of multiple objective performance such as distribution network protection speed, selectivity, and economy.

[0004] Currently, research and application of site selection strategies for hierarchical protection in distribution networks in China are very limited. Existing standards define protection configuration and setting principles for 3-10 kV distribution networks, but neither standard provides systematic requirements or methods for the specific location selection of protection devices. Prior art proposes quantifying hierarchical protection for distribution networks, dividing the protection scheme into three levels of protection, but lacks a method for categorizing the different levels, nor does it propose a site selection strategy for hierarchical protection of distribution lines. Prior art proposes branch-trunk coordination rules based on hierarchical protection concepts, but fails to quantify global constraints on the location and number of protection devices. Prior art proposes a configuration scheme for circuit breakers on distribution network trunk lines, optimizing the locations of the trunk line circuit breakers based on their inherent characteristics and positional relationships. However, it fails to consider the protection configuration and site selection of branch lines during the planning phase. Prior art uses a particle swarm optimization algorithm to optimize protection operation time, but this model simplifies the protection hierarchy of distribution line lines, limiting the scheme's applicability. Existing technologies have proposed an optimized gray wolf algorithm to improve distribution network hierarchical protection. However, its excessive number of pre-set parameters reduces the algorithm's simplicity and significantly reduces its computational speed. Furthermore, existing technologies involve multi-objective optimization algorithms, but their constraints fail to address practical factors such as the lower limit of protection locations and the number of protection configurations. They also lack research on protection point selection, and the algorithms are prone to falling into local optimality. Overall, existing methods and standards still need improvement, their theoretical models and research scope are limited, and their algorithmic applicability is insufficient. Summary of the Invention

[0005] In order to solve the deficiencies in the prior art, the present invention provides a method and system for optimizing the configuration of hierarchical protection points in a distribution network.

[0006] The present invention adopts the following technical solutions.

[0007] The first aspect of the present invention provides a method for optimizing the configuration of points for hierarchical protection of a distribution network, comprising the following contents:

[0008] Setting multiple objective functions and multiple constraints for protection point selection optimization, wherein the protection point selection optimization is to optimize the number and location of protection points; establishing a constraint violation degree calculation formula based on all the constraints;

[0009] The distribution network parameters are obtained and, based on these parameters, a particle swarm optimization algorithm that introduces a three-dimensional vector rotation matrix and a dynamic perturbation vector is used to iteratively optimize multiple objective functions. During the optimization process, particles are screened based on the degree of constraint violation, and the particle priority is calculated based on the constraint violation degree and the congestion distance.

[0010] A set number of protection points are randomly selected from the multiple protection points found, and the final protection points are obtained through clustering and screening. Preferably, the constraints include positional relationship constraints between hierarchical protections, hierarchical protection configuration quantity constraints, trunk line protection position constraints, and time coordination constraint functions; the positional relationship constraints between hierarchical protections are that the trunk line protection position is upstream of the corresponding branch protection in the same level protection; the hierarchical protection configuration quantity constraints are that the number of trunk line protections and the number of branch line protections are within a set range; the trunk line protection position constraints are that the second-level protection position of the distribution network in the trunk line satisfies the sensitivity of the substation I protection setting value of the distribution network line reaches a set sensitivity threshold, and the third-level protection position of the distribution network in the trunk line satisfies the sensitivity of the substation II protection setting value of the distribution network line reaches a set sensitivity threshold; the time coordination constraint function is that the action time between the same protection sections I, II, and III must maintain that the action time of section I is greater than the action time of section II, the action time of section II is greater than the action time of section III, and the difference in action time between different protections is within the set corresponding time difference threshold.

[0011] Preferably, the objective function includes an objective function of total loss load, an objective function of important loss load and an objective function of the number of households with load loss;

[0012] The objective function of the total loss load is to minimize the sum of the load losses caused by power outages during the short circuit period when the distribution network has a set number of short circuits.

[0013] The objective function of important loss load is to set the importance weight according to the importance of load short circuit, and minimize the short circuit period when the distribution network has a set number of short circuits. The load loss caused by power outage is weighted summed according to the importance weight.

[0014] The objective function of the number of households with load loss sets a comprehensive impact weight, and calculates the total number of users in the distribution network affected by power outages and the weighted summation result based on the comprehensive impact weight.

[0015] Preferably, the comprehensive impact weight is specifically:

[0016] Obtain all affected substations and their corresponding user types in the distribution network, set a basic weight for each user type, set a power outage time threshold for the affected substations in the distribution network to represent the power consumption sensitivity of the substation, and obtain the scope of the chain reaction when the power outage occurs in the affected substations in the network;

[0017] Comprehensive impact weight , the formula is:

[0018]

[0019] Where, For the gThe basic weights set for the user types in the affected areas of the distribution network; For the g The power outage time threshold set for each affected area in the distribution network; For the g The scope of chain reaction when power outage occurs in the affected substations in the distribution network.

[0020] Preferably, the constraint violation degree calculation formula established based on all constraint conditions is specifically:

[0021]

[0022] Where, For particles x The degree of constraint violation; K is the number of constraints; is the number of equality constraints; For the u The degree of violation of the constraint conditions; is the number of all particles in the current iteration during the optimization process. u The maximum value among the violation degrees of the constraints; 、 All are u Constraints, when u When the constraint condition is an equality constraint, , when it is an inequality constraint, ; is the set tolerance factor.

[0023] Preferably, the iterative optimization of the multiple objective functions using a particle swarm algorithm that introduces a three-dimensional vector rotation matrix and a dynamic perturbation vector is performed as follows:

[0024] After setting the initial particle swarm size, number of short circuits, and maximum number of iterations, randomly generate the particle swarm;

[0025] Calculate the constraint violation degree of all particles and delete particles below the set degree threshold; combine the three-dimensional vector rotation matrix and the dynamic perturbation vector to update the particle position; calculate the particle priority of each particle based on the constraint violation degree and crowding distance, and update the leader of each particle according to the particle priority; repeat the above steps starting from calculating the constraint violation degree of all particles until the maximum number of iterations is reached.

[0026] Preferably, the three-dimensional vector rotation matrix and the dynamic disturbance vector are combined to update the particle position, specifically:

[0027] Where, for t +1 iteration generates new particle positions; for t The search center of the t iteration is equal to the sum of the global leader and the individual leader of the t iteration; is the search range coefficient, , rand1 is a random number, , is the maximum number of iterations; is the three-dimensional vector rotation matrix; 、 are the global guide and individual guide of t iterations respectively; is the disturbance direction coefficient, , rand2 is a random number, ; is a unit random vector consisting of three elements whose sum of squares is 1.

[0028] Preferably, the particle priority of each particle is calculated by combining the constraint violation degree and the crowding distance, specifically:

[0029] After each iteration, the constraint violation degree and crowding distance of each particle are calculated; when the constraint violation degree is 0, the priority is equal to the crowding distance; when the constraint violation degree is not 0, the priority is equal to the crowding distance multiplied by the constraint violation degree.

[0030] Preferably, the updating of the leader of each particle according to the particle priority is specifically as follows:

[0031] Sort the particles by priority from largest to smallest, select the first set number of particles, and update the individual guides of these particles;

[0032] A particle position is selected to be added to a reserve set, wherein the reserve set includes a feasible set and an infeasible set, and the initial infeasible reserve set and the feasible reserve set are set to an empty set and all zeros; particles that meet all constraints are added to the feasible set, and particles that violate one or more constraints are added to the infeasible set; a first probability and a second probability are set, wherein the first probability is greater than the second probability, and both the first and second probabilities are greater than or equal to 0 and less than or equal to 1, a quotient of the current number of iterations and the maximum number of iterations is calculated and multiplied by the second probability, and the product is subtracted from the first probability as a third probability; and a particle position is selected from the infeasible reserve set and the feasible reserve set as an updated global leader using the third probability and 1 minus the third probability as the selection probabilities, respectively.

[0033] Preferably, the final protection points are obtained by clustering and screening, specifically:

[0034] Calculate all objective function values ​​of a set number of randomly selected protection points respectively, use all objective function values ​​as the coordinates of the corresponding protection points, cluster all protection points according to their coordinates, and the number of clusters is the same as the number of objective functions. Select the protection points with the closest Euclidean distance to each cluster center respectively, and normalize them. Take the weighted sum of all objective function values ​​of each normalized protection point as the expectation of the corresponding protection point, with the weight being the set value, and select the protection point with the smallest expectation as the final protection point.

[0035] The second aspect of the present invention proposes a distribution network hierarchical protection point selection optimization configuration system based on the method described in the first aspect of the present invention, including an optimization target and constraint condition construction module, a constraint violation degree establishment module, an optimization module, and an optimization module, specifically:

[0036] Optimization objective and constraint condition construction module: sets multiple objective functions and multiple constraints for protection point selection optimization, wherein the protection point selection optimization is to optimize the number and location of protection points;

[0037] Constraint violation degree establishment module: establishes a constraint violation degree calculation formula based on all constraint conditions;

[0038] Optimization module: This module obtains distribution network parameters and uses a particle swarm algorithm that introduces a three-dimensional vector rotation matrix and a dynamic perturbation vector to iteratively optimize multiple objective functions based on these parameters. During the optimization process, particles are screened based on the degree of constraint violation and particle priority is calculated based on the degree of constraint violation and congestion distance.

[0039] Optimization module: randomly selects a set number of protection points from the multiple protection points found, and obtains the final protection points through clustering and screening.

[0040] The beneficial effects of the present invention lie in that, compared to existing technologies, it constructs a multidimensional optimization model based on positional constraints, protection quantity limits, and trunk line protection configuration location and timing coordination as optimization constraints, with minimizing multiple loss loads as the objective function. The protection quantity limits incorporate length-correlation constraints, innovatively determine the load importance and load loss coefficient of distributed generation (DGs) to quantify the significant loss load costs, and assign comprehensive impact weights to focus on user-side impacts, quantifying the differential impact of power outages on different types of users. The present invention selects update leaders based on a constraint violation degree judgment formula and congestion distance. When updating particle positions, the present invention introduces a three-dimensional vector rotation matrix and a dynamic perturbation vector, enabling particles to perform multi-angle rotation searches within a spatial plane determined by individual and global optima. The search range is adaptively adjusted based on the number of iterations. Without requiring pre-set parameters such as inertia weights and learning factors, the present invention overcomes the planar limitations of traditional linear search and enables three-dimensional exploration of the multidimensional objective space (i.e., multiple objective functions) for distribution network protection point selection, significantly improving the distribution uniformity of the solution set and global search capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a flow chart of the method of this embodiment;

[0042] Figure 2 This is the distribution network node system of this embodiment;

[0043] Figure 3 The distribution of the optimal solutions of the protection point selection optimization model of this embodiment;

[0044] Figure 4 This is the clustering result of this embodiment. DETAILED DESCRIPTION

[0045] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. The embodiments described in this application are only part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0046] like Figure 1 As shown, embodiment 1 of the present invention proposes a method for optimizing the configuration of points for hierarchical protection of a distribution network, including:

[0047] Multiple objective functions and multiple constraints for protection point selection optimization are set, wherein the protection point selection optimization is to optimize the number and location of protection points.

[0048] Specifically, the matrix consisting of the protected position and the protected quantity is used as the variable for subsequent optimization.y :

[0049]

[0050]

[0051]

[0052] in, To protect the position matrix, To protect the number matrix; 、 They are the positions of main line protection and branch line protection respectively. i Representative for the distribution network i Level protection, j Indicates the position in the main line or branch protection of the same level; 、 are the protection quantities of the trunk line and branch line respectively, and T is the transposition symbol.

[0053] In the preferred scheme of this embodiment, the constraints include position relationship constraints between hierarchical protections, hierarchical protection configuration quantity constraints, main line protection position constraints and time coordination constraint functions; the position relationship constraints between hierarchical protections are that the main line protection position in the same level protection is upstream of the corresponding branch protection; the hierarchical protection configuration quantity constraints are that the number of main line protections and the number of branch line protections are between the set ranges; the main line protection position constraints are that the second-level protection position of the distribution network in the main line satisfies the sensitivity of the substation I section protection constant of the distribution network line reaches the set sensitivity threshold, and the third-level protection position of the distribution network in the main line satisfies the sensitivity of the substation II section protection constant of the distribution network line reaches the set sensitivity threshold; the time coordination constraint function is that the action time between the same protection sections I, II and III must keep the action time of section I greater than the action time of section II, the action time of section II greater than the action time of section III, and the difference in action time between different protections is within the set corresponding time difference threshold.

[0054] The positional relationship constraint between hierarchical protections is: the trunk line protection position in the same level protection is upstream of the corresponding branch protection. The formula is expressed as: ;in, Indicates any position in the main line or branch protection at the same level;

[0055] The number of hierarchical protection configuration constraints is: the number of trunk line protection and the number of branch line protection are within the set range. The formula is:

[0056]

[0057] in, To protect the total number; The total number of main lines; is the total number of branches; For the a Number of protection lines in the trunk line; For the b Number of protected branches; 、 Respectively a The length of the main line, b The length of the branch; The maximum length threshold for setting no protection.

[0058] The trunk line protection position constraints are as follows: the second-level protection position of the distribution network in the trunk line satisfies the sensitivity of the substation I section protection setting value of the distribution network line to reach the set sensitivity threshold, and the third-level protection position of the distribution network in the trunk line satisfies the sensitivity of the substation II section protection setting value of the distribution network line to reach the set sensitivity threshold;

[0059] Specifically, the sensitivity threshold is set to 1.3, and the I-stage protection constant is calculated by the two-phase short-circuit current at the distribution network trunk line; at the same time, the I-stage constant is determined according to the maximum short-circuit current at the installation location of the main line protection device. Based on this, the constant is divided by the reliability coefficient to obtain the three-phase short-circuit current value of the secondary protection selection point of the main line protection under the maximum operating mode; from this, the total system impedance under the maximum operating mode can be calculated. Combined with the line impedance at the protection selection point, the lower limit of the selection point closest to the substation end of the secondary main line protection can be determined; the final obtained second-level protection position of the distribution network in the main line The constraint formula is:

[0060]

[0061] in, 、 are the system resistance and reactance under maximum operation mode respectively; 、 、 are the unit impedance, resistance and reactance of the distribution network line conductor respectively; is the system impedance under the minimum operating mode;

[0062] Replace the protection setting of section I with the protection setting of section II, and finally obtain the third-level protection position of the distribution network in the trunk line. The constraint formula is:

[0063]

[0064] The time coordination constraint function is to meet the coordination of the action time between the I, II, and III sections of the same protection, and the coordination of the action time between different protections. The formula is:

[0065]

[0066] Where, 、 、 Respectively o The action time of protection II and III sections; 、 、 Respectively The action time of protection II and III sections; The protection is o A protected subordinate.

[0067] It should be noted that the upper protection current segment II can be coordinated with the lower protection current segment I or II, and the upper protection current segment III can be coordinated with the lower protection current segment III. The action time of the coordinated segments (i.e. the difference between the two) must be greater than the set time difference threshold. , this embodiment Set to 0.2s; The overcurrent protection action time limit is set to 0.7s in this embodiment.

[0068] The objective functions include a total loss load objective function, a major loss load objective function, and a load loss household objective function;

[0069] In the new active distribution network, distributed power generation can maintain the supply of protection cut-off loads and reduce the loss of loads. When a fault occurs, DG will give priority to supplying power to the load, reducing the "forced load shedding amount". In addition to the main station for fault identification and command issuance, the distribution network load shedding control system also includes multiple control substations and a large number of load terminals. The control substation can be abstracted as the load user of all loads within its controllable range, and each load terminal is part of the power consumption of the load user. Through the load loss coefficient Quantify the DG's compensation effect on the total load loss of each control substation - if the DG can fully cover the load power, the load does not need to be cut off; if the DG power is insufficient, the excess load is cut off to ensure system stability, and the covered part is regarded as loss The bigger it is, the more you lose.

[0070] Total loss load objective function In order to minimize the load summation of power outage losses during the short circuit period when the distribution network has a set number of short circuits, the formula is:

[0071]

[0072] in, For load Power; is the load loss coefficient; After the protection action, the load power provided by the distributed power supply is n is the total load, N The number of short circuits is set.

[0073] Important loss load objective function In order to set the importance weight according to the importance of load short circuit, the weighted sum of the load losses during the short circuit period when the distribution network has a set number of short circuits is minimized. The formula is:

[0074]

[0075] in, For the setting q Secondary short circuit load The importance of q Secondary short circuit load Set the grading and classification of For load Power; is the load loss coefficient; After the protection action, the load power provided by the distributed power supply is n is the total load, N The number of short circuits is set.

[0076] Load loss household objective function To set the comprehensive impact weight, calculate the total number of users in the distribution network affected by power outages and perform a weighted summation based on the comprehensive impact weight.

[0077] Set basic weights based on user types, and construct comprehensive impact weights by integrating basic weights of user types, sensitivity of power outage duration, and chain effects. , the formula is:

[0078]

[0079] Where, For the g The basic weights set for the user types in the affected areas of the distribution network; For the g The power outage time threshold set for each affected area in the distribution network; Reflects the sensitivity of power outage duration, The smaller the value, the more sensitive the load user is, and the larger the value is, such as in hospitals. Smaller, Larger, increase its comprehensive weight; For the g The larger the scope of the chain reaction when a power outage occurs in the affected substations in the distribution network, the higher the comprehensive weight. For example, if the power outage in a commercial complex has a large chain reaction, the comprehensive impact weight will be increased.

[0080] The actual distribution network is based on the control range of the substation, which is the smallest power supply unit. One substation is powered by one distribution transformer. Each substation corresponds to a fixed user list. The objective function of the load loss number is The formula is:

[0081]

[0082] Where: G is the total number of affected substations in the distribution network; For the g The power outage status of the affected substation in the distribution network, 1 for power outage and 0 for no power outage; For the g The number of registered users in the affected areas of the distribution network is extracted from the marketing system.

[0083] Establish a constraint violation degree calculation formula based on all constraint conditions;

[0084] In the preferred solution of this embodiment, the constraint violation degree calculation formula is established based on all constraint conditions, specifically:

[0085]

[0086] Where, For particles x The degree of constraint violation; K is the number of constraints; is the number of equality constraints; For the u The degree of violation of the constraint conditions; is the number of all particles in the current iteration during the optimization process. u The maximum value among the violation degrees of the constraints; 、 All are u Constraints, when u When the constraint condition is an equality constraint, , when it is an inequality constraint, ; is the set tolerance factor.

[0087] The distribution network parameters are obtained, and based on these parameters, a particle swarm algorithm that introduces a three-dimensional vector rotation matrix and a dynamic perturbation vector is used to iteratively optimize multiple objective functions. During the optimization process, particles are screened according to the degree of constraint violation, and the particle priority is calculated based on the constraint violation degree and the congestion distance.

[0088] Specifically, this embodiment is as follows Figure 2 The IEEE 33-node system shown in the figure is used as an example; the rated voltage level of the system is 10.5 kV. The parameters set in the initial stage of the algorithm are shown in Table 1. The impedance of the distribution network system under the maximum and minimum operating modes is , ,The distribution network line parameters and load parameters are shown in Table 1 and Table 2.

[0089] Table 1 Distribution network line parameters

[0090]

[0091] Table 2 Distribution network load parameters

[0092]

[0093] Table 2 Is an imaginary number.

[0094] In the preferred solution of this embodiment, the particle swarm algorithm that introduces a three-dimensional vector rotation matrix and a dynamic perturbation vector is used to iteratively optimize the multiple objective functions, specifically:

[0095] Set the initial particle swarm size and number of short circuits N , maximum number of iterations Then randomly generate particle swarm;

[0096] Calculate the constraint violation degree of all particles and delete particles below the set degree threshold; combine the three-dimensional vector rotation matrix and the dynamic perturbation vector to update the particle position; calculate the particle priority of each particle based on the constraint violation degree and crowding distance, and update the leader of each particle according to the particle priority; repeat the above steps starting from calculating the constraint violation degree of all particles until the maximum number of iterations is reached.

[0097] It should be noted that the particle position is an optimized variable y Each particle is a set of multiple candidate solutions. The guides include the global guide and the individual guide. The global guide is the optimal solution found by the entire particle swarm from the beginning of the iteration to the current iteration number, and the individual guide is the optimal solution encountered by each particle in each update of itself.

[0098] Specifically, the initial simulation parameters set in this embodiment include the initial particle swarm size, the number of short circuits,N , maximum number of iterations It also includes parameters such as reserve set capacity and non-reserve set capacity, as shown in Table 3.

[0099] Table 3 Simulation initial parameters

[0100]

[0101] In the preferred solution of this embodiment, the three-dimensional vector rotation matrix and the dynamic disturbance vector are combined to update the particle position, specifically:

[0102] Where, for t +1 iteration generates new particle positions; for t The search center of the t iteration is equal to the sum of the global leader and the individual leader of the t iteration; is the search range coefficient, , rand1 is a random number, , is the maximum number of iterations; is the three-dimensional vector rotation matrix; 、 are the global guide and individual guide of t iterations respectively; is the disturbance direction coefficient, , rand2 is a random number, ; is a unit random vector consisting of three elements, and the sum of the squares of the three elements is 1, that is ,and , are the three elements of the vector respectively.

[0103] In a preferred solution of this embodiment, the particle priority of each particle is calculated by combining the constraint violation degree and the crowding distance, specifically:

[0104] After each iteration, the constraint violation degree and crowding distance of each particle are calculated; when the constraint violation degree is 0, the priority is equal to the crowding distance; when the constraint violation degree is not 0, the priority is equal to the crowding distance multiplied by the constraint violation degree.

[0105] Specifically, the crowding distance calculation formula is:

[0106]

[0107] Where, Represents the current solution of the particle The crowding distance, For the Solution No. The value of the objective function, To refer to Solution No. m The value of the objective function; For all solutions of this particle m The maximum value of the objective function, is the number of objective functions; all objective functions of each solution are formed into coordinates, 、 for peace The two solutions with the closest Euclidean distance of corresponding coordinates.

[0108] In the preferred solution of this embodiment, the updating of the leader of each particle according to the particle priority is specifically as follows:

[0109] Sort the particles by priority from largest to smallest, select the first set number of particles, and update the individual guides of these particles;

[0110] Select particle positions to be added to a reserve set, where the reserve set includes a feasible set and an infeasible set. Initially, the infeasible reserve set and the feasible reserve set are set to an empty set and all zeros. Particles that satisfy all constraints are added to the feasible set, and particles that violate one or more constraints are added to the infeasible set. Set a first probability and a second probability, where the first probability is greater than the second probability, and both the first and second probabilities are greater than or equal to 0 and less than or equal to 1. Calculate the quotient of the current number of iterations and the maximum number of iterations multiplied by the second probability, and use the first probability minus the product as the third probability. Use the third probability and 1 minus the third probability as the selection probabilities to select particle positions from the infeasible reserve set and the feasible reserve set as the updated global leader. The formula is:

[0111]

[0112] in, 、 、 are the first, second, and third probabilities respectively.

[0113] Specifically, Figure 3 This is the distribution of the optimal solutions for the protection point selection optimization model, where each particle represents a set of protection point selection solutions. As can be seen from the figure, the resulting solution set is relatively evenly distributed and has good diversity, demonstrating the strong solution-solving capabilities of the proposed point selection model.

[0114] In a preferred solution of this embodiment, a set number of protection points are randomly selected from the multiple protection points found optimally, and the final protection points are obtained through clustering and screening.

[0115] The final protection points are obtained through clustering and screening, specifically:

[0116] Calculate all objective function values ​​of a set number of randomly selected protection points respectively, use all objective function values ​​as the coordinates of the corresponding protection points, cluster all protection points according to their coordinates, and the number of clusters is the same as the number of objective functions. Select the protection points with the closest Euclidean distance to each cluster center respectively, and normalize them. Take the weighted sum of all objective function values ​​of each normalized protection point as the expectation of the corresponding protection point, with the weight being the set value, and select the protection point with the smallest expectation as the final protection point.

[0117] In the preferred solution of this embodiment, clustering first randomly sets the cluster center, uses the Euclidean distance method to calculate the distance of each coordinate to all cluster centers, and assigns it to the Euclidean closest cluster, takes the mean of all coordinates in the cluster as the center of the new corresponding cluster, and repeats the above steps until the cluster center no longer changes.

[0118] In the preferred solution of this embodiment, all solutions of the cluster center in the optimization process are obtained, and the maximum and minimum values ​​of all their objective functions are obtained; when the objective function value of the optimal solution after the cluster center optimization is less than or equal to the minimum value of the corresponding objective function, it is normalized to 1; if it is greater than or equal to the corresponding maximum value, it is normalized to 0; otherwise, the normalization is equal to the difference between the objective function values ​​of the optimal solution and the minimum value divided by the difference between the maximum value and the minimum value.

[0119] The clustering results are as follows Figure 4 As shown, by comparing the data of the traditional distribution network protection point selection configuration scheme shown in Table 4 with the three sets of protection point selection obtained after clustering shown in Tables 5, 6, and 7, it can be seen that the distribution network hierarchical protection point selection optimization method proposed in this embodiment can significantly reduce the total action time of each protection, the loss load, and the total investment in protection equipment during distribution network faults. The total load loss is reduced by up to 9.21%, the important load loss is reduced by up to 7.32%, and the number of households with lost load is reduced by up to 18.12%. The reliability and economy of distribution network protection are improved, and the overall applicability of the global optimal solution for protection point selection is proved.

[0120] Table 4. Configuration scheme of protection point selection in traditional distribution network

[0121]

[0122] Table 5: Protection point selection configuration scheme for the first set

[0123]

[0124] Table 6 Protection point selection configuration scheme for set II

[0125]

[0126] Table 7 Protection point selection scheme for set III

[0127]

[0128] Table 8 Normalization parameters of each objective function

[0129]

[0130] Table 9 Normalized expectation of protection site selection scheme

[0131]

[0132] The normalized parameters of each objective function are calculated as shown in Table 8. According to the data shown in Table 9, the protection point selection scheme II with the minimum expectation is finally selected as the optimal scheme for the hierarchical protection point selection of the distribution network.

[0133] The present disclosure may be a system, method and / or computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0134] Finally, it should be noted that 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 above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A method for optimizing the configuration of points for hierarchical protection in a distribution network, characterized in that: Includes the following: Setting multiple objective functions and multiple constraints for protection point selection optimization, wherein the protection point selection optimization is to optimize the number and location of protection points; establishing a constraint violation degree calculation formula based on all the constraints; The objective functions include a total loss load objective function, a major loss load objective function, and a load loss household objective function; The objective function of the total loss load is to minimize the sum of the load losses caused by power outages during the short circuit period when the distribution network has a set number of short circuits. The objective function of important loss load is to set the importance weight according to the importance of load short circuit, and minimize the short circuit period when the distribution network has a set number of short circuits. The load loss caused by power outage is weighted summed according to the importance weight. The objective function of the number of households with load loss sets a comprehensive impact weight, and calculates the total number of users in the distribution network affected by power outages and the weighted sum of the comprehensive impact weights; The constraint violation degree calculation formula is established based on all constraint conditions, specifically: Where, For particles x The degree of constraint violation; K is the number of constraints; is the number of equality constraints; For the u The degree of violation of the constraint conditions; is the number of all particles in the current iteration during the optimization process. u The maximum value among the violation degrees of the constraints; 、 All are u Constraints, when u When the constraint condition is an equality constraint, , when it is an inequality constraint, ; is the set tolerance factor; The distribution network parameters are obtained and, based on these parameters, a particle swarm optimization algorithm that introduces a three-dimensional vector rotation matrix and a dynamic perturbation vector is used to iteratively optimize multiple objective functions. During the optimization process, particles are screened based on the degree of constraint violation, and the particle priority is calculated based on the constraint violation degree and the congestion distance. The particle swarm algorithm that introduces a three-dimensional vector rotation matrix and a dynamic perturbation vector is used to iteratively optimize multiple objective functions, specifically: After setting the initial particle swarm size, number of short circuits, and maximum number of iterations, randomly generate the particle swarm; Calculate the constraint violation degree of all particles and delete particles below the set degree threshold; combine the three-dimensional vector rotation matrix and the dynamic perturbation vector to update the particle position; calculate the particle priority of each particle based on the constraint violation degree and crowding distance, and update the leader of each particle according to the particle priority; repeat the above steps starting from calculating the constraint violation degree of all particles until the maximum number of iterations is reached; A set number of protection points are randomly selected from the multiple protection points found, and the final protection points are obtained through clustering and screening.

2. The method for optimizing the configuration of points for hierarchical protection in a distribution network according to claim 1, characterized in that: The constraints include positional relationship constraints between hierarchical protections, number constraints for hierarchical protection configurations, position constraints for trunk line protections, and time coordination constraint functions; the positional relationship constraints between hierarchical protections are that the trunk line protection position in the same level protection is upstream of the corresponding branch protection; the number constraints for hierarchical protection configurations are that the number of trunk line protections and the number of branch line protections are within a set range; the trunk line protection position constraints are that the second-level protection position of the distribution network in the trunk line satisfies the sensitivity of the substation I section protection constant of the distribution network line reaches the set sensitivity threshold, and the third-level protection position of the distribution network in the trunk line satisfies the sensitivity of the substation II section protection constant of the distribution network line reaches the set sensitivity threshold; the time coordination constraint function is that the action time between sections I, II, and III of the same protection must keep the action time of section I greater than the action time of section II, the action time of section II greater than the action time of section III, and the difference in action time between different protections is within the set corresponding time difference threshold.

3. The method for optimizing the configuration of points for hierarchical protection in a distribution network according to claim 1, characterized in that: The comprehensive impact weights are specifically: Obtain all affected substations and their corresponding user types in the distribution network, set a basic weight for each user type, set a power outage time threshold for the affected substations in the distribution network to represent the power consumption sensitivity of the substation, and obtain the scope of the chain reaction when the power outage occurs in the affected substations in the network; Comprehensive impact weight , the formula is: Where, For the g The basic weights set for the user types in the affected areas of the distribution network; For the g The power outage time threshold set for each affected area in the distribution network; For the g The scope of chain reaction when power outage occurs in the affected substations in the distribution network.

4. The method for optimizing the configuration of points for hierarchical protection in a distribution network according to claim 1, characterized in that: The particle position is updated by combining the three-dimensional vector rotation matrix and the dynamic disturbance vector, specifically: Where, for t +1 iteration generates new particle positions; for t The search center of the t iteration is equal to the sum of the global leader and the individual leader of the t iteration; is the search range coefficient, , rand1 is a random number, , is the maximum number of iterations; is the three-dimensional vector rotation matrix; 、 are the global guide and individual guide of t iterations respectively; is the disturbance direction coefficient, , rand2 is a random number, ; is a unit random vector consisting of three elements, and the sum of the squares of the three elements is 1.

5. The method for optimizing the configuration of points for hierarchical protection in a distribution network according to claim 1, characterized in that: The particle priority of each particle is calculated by combining the constraint violation degree and the crowding distance, specifically: After each iteration, the constraint violation degree and crowding distance of each particle are calculated; when the constraint violation degree is 0, the priority is equal to the crowding distance; when the constraint violation degree is not 0, the priority is equal to the crowding distance multiplied by the constraint violation degree.

6. The method for optimizing the configuration of points for hierarchical protection in a distribution network according to claim 1, characterized in that: The process of updating the leader of each particle according to the particle priority is as follows: Sort the particles by priority from largest to smallest, select the first set number of particles, and update the individual guides of these particles; A particle position is selected to be added to a reserve set, wherein the reserve set includes a feasible set and an infeasible set, and the initial infeasible reserve set and the feasible reserve set are set to an empty set and all zeros; particles that meet all constraints are added to the feasible set, and particles that violate one or more constraints are added to the infeasible set; a first probability and a second probability are set, wherein the first probability is greater than the second probability, and both the first and second probabilities are greater than or equal to 0 and less than or equal to 1, a quotient of the current number of iterations and the maximum number of iterations is calculated and multiplied by the second probability, and the product is subtracted from the first probability as a third probability; and a particle position is selected from the infeasible reserve set and the feasible reserve set as an updated global leader using the third probability and 1 minus the third probability as the selection probabilities, respectively.

7. The method for optimizing the configuration of points for hierarchical protection in a distribution network according to claim 1, characterized in that: The final protection points are obtained through clustering and screening, specifically: Calculate all objective function values ​​of a set number of randomly selected protection points respectively, use all objective function values ​​as the coordinates of the corresponding protection points, cluster all protection points according to their coordinates, and the number of clusters is the same as the number of objective functions. Select the protection points with the closest Euclidean distance to each cluster center respectively, and normalize them. Take the weighted sum of all objective function values ​​of each normalized protection point as the expectation of the corresponding protection point, with the weight being the set value, and select the protection point with the smallest expectation as the final protection point.

8. A distribution network hierarchical protection point selection optimization configuration system based on the method according to any one of claims 1 to 7, comprising an optimization target and constraint condition construction module, a constraint violation degree establishment module, an optimization module, and an optimization module, characterized in that: Optimization objective and constraint condition construction module: sets multiple objective functions and multiple constraints for protection point selection optimization, wherein the protection point selection optimization is to optimize the number and location of protection points; Constraint violation degree establishment module: establishes a constraint violation degree calculation formula based on all constraint conditions; Optimization module: This module obtains distribution network parameters and uses a particle swarm algorithm that introduces a three-dimensional vector rotation matrix and a dynamic perturbation vector to iteratively optimize multiple objective functions based on these parameters. During the optimization process, particles are screened based on the degree of constraint violation and particle priority is calculated based on the degree of constraint violation and congestion distance. Optimization module: randomly selects a set number of protection points from the multiple protection points found, and obtains the final protection points through clustering and screening.

Images