Section switch layout planning method and device of power distribution network, equipment and medium

Abstract

The invention discloses a section switch layout planning method and device for a power distribution network, equipment and a medium, and belongs to the field of section switch layout of the power distribution network, and the method comprises the steps: after obtaining a plurality of different switch point distribution schemes, calling an MIOA algorithm to carry out the iteration of the plurality of switch point distribution schemes, and obtaining a local scheme set; calling an SDT algorithm to perform search iteration on the schemes of the local scheme set to obtain a convergence scheme set; calculating the total cost of each scheme of the convergence scheme set based on a target function of a preset target model to obtain a plurality of scheme cost values; and screening a target cost value from the plurality of scheme cost values, determining a scheme corresponding to the target cost value as a target planning scheme, and performing layout processing on the section switch of the power distribution network based on the target planning scheme. The scheme meeting the requirements of the power distribution network can be screened, the deviation between the section switch layout and the reality is reduced, the configuration precision is improved, and the layout cost of the section switch can be reduced.

Description

Technical Field

[0001] This invention relates to the technical field of distribution network sectional switch planning, and in particular to a method, apparatus, equipment and medium for sectional switch layout planning in distribution networks. Background Technology

[0002] In the planning and development of power systems, the distribution network, as a crucial link connecting the transmission system and end users, plays a vital role in ensuring power supply reliability and improving grid operation efficiency by enhancing its automation and intelligence levels. Statistics show that approximately 70% of customer power outages originate from distribution network faults, and the frequency and duration of these outages are important reliability indicators for measuring distribution network performance.

[0003] To improve power supply availability and shorten fault recovery time, an effective approach is to rationally configure sectionalizing switches. One commonly used method for planning sectionalizing switches is to evenly divide the number of users and loads, using this as a basis for laying out automated switches. This method is simple to operate and easy to implement, and can meet basic power grid operation requirements to a certain extent.

[0004] However, the currently used methods have the following technical problems: different users have the same electricity demand and load, and in the event of a distribution network fault, independent power supply areas (i.e., islands) are formed through distributed generation (DG), requiring DG to supply power to different users, thus causing the power supply structure of the distribution network to change constantly. Existing methods are difficult to adapt to users with different needs and the daytime variable distribution network structure, resulting in a large deviation between the configured sectionalizing switch layout and the actual situation, and requiring high configuration accuracy. This not only fails to meet current usage requirements but also increases the layout cost of sectionalizing switches. Summary of the Invention

[0005] This invention provides a method, apparatus, equipment, and medium for planning the layout of sectional switches in a power distribution network, which can solve the technical problems of large deviations between the planned sectional switches and the actual situation, and low planning and matching accuracy in the prior art.

[0006] A first aspect of this invention provides a method for planning the layout of sectionalizing switches in a power distribution network, the method comprising: After obtaining multiple different switch placement schemes, the MIGOA algorithm is called to iterate the multiple switch placement schemes to obtain a local scheme set, wherein each of the switch placement schemes is a layout planning scheme for segmented switches in the distribution network. The SDT algorithm is invoked to search and iterate over the solutions in the local solution set to obtain a convergent solution set; The total cost of each scheme in the convergence scheme set is calculated based on the objective function of the preset target model to obtain multiple scheme cost values. The objective function of the preset target model is a function constructed from the downtime of load points inside and outside the island and the cost of sectionalizing the switch. Select a target cost value from multiple proposed cost values, determine the proposed solution corresponding to the target cost value as the target planning solution, and perform layout processing on the sectional switches of the distribution network based on the target planning solution.

[0007] This invention can obtain multiple different switch layout schemes. By using the MIGOA algorithm and the SDT algorithm, several switch layout schemes that meet the needs of the distribution network can be selected from multiple schemes. The cost of each scheme is calculated by constructing an objective function based on cost, and then the scheme is selected based on the cost. This not only selects schemes that meet the needs of the distribution network, reduces the deviation between the segmented switch layout and the actual situation, and improves the configuration accuracy, but also reduces the layout cost of segmented switches.

[0008] In conjunction with the first aspect, in one implementation, after obtaining multiple different switch placement schemes, the MIGOA algorithm is invoked to iterate through the multiple switch placement schemes to obtain a local scheme set, including: Multiple different switch placement schemes are obtained, and the multiple switch placement schemes are divided into N preprocessed subgroups, wherein each subgroup contains several individuals, and each individual is a switch placement scheme. Calculate the fitness of each individual in each of the subpopulations, and migrate individuals within the multiple subpopulations based on the fitness until the individuals in each subpopulation converge, thus obtaining a set of local solutions.

[0009] In conjunction with the first aspect, in one implementation, obtaining multiple different switch placement schemes and dividing the multiple switch placement schemes into N preprocessed subgroups includes: Multiple different switch placement schemes are obtained, and the multiple switch placement schemes are divided into N initial populations, wherein each initial population includes multiple individuals; Each individual in the initial population is preprocessed to obtain a subpopulation, wherein the preprocessing is the process of removing similar individuals from each of the initial populations.

[0010] In conjunction with the first aspect, in one implementation, the preprocessing of individuals within each initial population to obtain a subpopulation includes: At least one individual is randomly selected from each of the initial populations as a reference. Calculate the Hamming distance between the reference body and each non-reference body in the initial population to obtain multiple individual distance values; Individuals whose body distance values ​​satisfy a preset distance value within the initial population are identified and removed to obtain a subpopulation.

[0011] In conjunction with the first aspect, in one implementation, the step of calculating the fitness of each individual in each of the subpopulations, and performing mutual migration of individuals within multiple subpopulations based on the fitness, until the individuals in each subpopulation converge to obtain a set of local solutions, includes: The fitness of each individual is calculated based on the shared function of each individual, and the mutation probability and crossover probability of each individual are calculated using the fitness. The roulette wheel method is used to randomly select corresponding individuals from each subpopulation as individuals to be processed. Based on the mutation probability and crossover probability of the individuals to be processed, the individuals to be processed in each subpopulation are migrated until the individuals in each subpopulation converge, thus obtaining a local scheme set.

[0012] In conjunction with the first aspect, in one implementation, the invocation of the SDT algorithm to search and iterate over the local solution set to obtain a convergent solution set includes: Select one subpopulation from the multiple subpopulations in the local solution set as the target population, wherein the target population is the subpopulation in the local solution set whose fitness satisfies a threshold. The gradient is calculated based on the target population using a differential algorithm, and a preset step size is determined, wherein the preset step size is a search step size calculated based on an adaptive binary search process; Starting with the target population, the optimal population is searched among multiple processing populations in the local scheme set according to the gradient and the preset step size until the difference between the optimal population and the target population meets the preset convergence threshold, and the individuals of the optimal population are used as the convergence scheme set.

[0013] In conjunction with the first aspect, in one implementation, the step of filtering a target cost value from multiple scheme cost values ​​and determining the scheme corresponding to the target cost value as the target planning scheme, and performing layout processing of the sectional switches of the distribution network based on the target planning scheme, includes: The minimum cost value among the multiple proposed cost values ​​is selected as the target cost value, and the proposed solution corresponding to the target cost value is determined as the target planning solution. Determine the location information of each segment switch within the target planning scheme, and perform layout processing of the segment switches of the distribution network based on the location information.

[0014] A second aspect of the present invention provides a sectional switch layout planning device for a power distribution network, the device comprising: The iteration module is used to call the MIGOA algorithm to iterate the multiple switch layout schemes after obtaining multiple different switch layout schemes, and obtain a local scheme set, wherein each switch layout scheme is a layout planning scheme of segmented switches in the distribution network. The search iteration module is used to call the SDT algorithm to search and iterate the solutions in the local solution set to obtain a converged solution set. The calculation module is used to calculate the total cost of each scheme in the convergence scheme set based on the objective function of the preset target model to obtain multiple scheme cost values. The objective function of the preset target model is a function constructed from the downtime of load points inside and outside the island and the cost of sectionalizing the switch. The layout planning module is used to filter target cost values ​​from multiple scheme cost values, determine the scheme corresponding to the target cost value as the target planning scheme, and perform layout processing on the sectional switches of the distribution network based on the target planning scheme.

[0015] Compared to existing technologies, the present invention provides a method, apparatus, equipment, and medium for segmented switch layout planning in a distribution network. Its advantages lie in the following: After obtaining multiple different switch placement schemes, the present invention uses the MIGOA algorithm to iterate through these schemes to obtain a local scheme set; then, the SDT algorithm is used to search and iterate through the schemes in the local scheme set to obtain a converged scheme set; based on the objective function of a preset target model, the total cost of each scheme in the converged scheme set is calculated to obtain multiple scheme cost values; a target planning scheme is selected from the multiple scheme cost values, and the segmented switches in the distribution network are then laid out based on the target planning scheme. The present invention can obtain multiple different switch placement schemes, and through the MIGOA and SDT algorithms, several switch placement schemes that meet the requirements of the distribution network can be selected. The cost of each scheme is calculated using a cost-based objective function, and the scheme is then selected based on its cost. This not only allows for the selection of schemes that meet the requirements of the distribution network, reducing the deviation between the segmented switch layout and the actual situation, and improving configuration accuracy, but also reduces the layout cost of the segmented switches. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a method for planning the layout of sectional switches in a power distribution network according to an embodiment of the present invention. Figure 2 This is an operation flowchart of a segmented switch layout planning method for a power distribution network provided by an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of a sectional switch layout planning device for a power distribution network provided in an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] In the planning and development of power systems, the distribution network, as a crucial link connecting the transmission system and end users, plays a vital role in ensuring power supply reliability and improving grid operation efficiency by enhancing its automation and intelligence levels. Statistics show that approximately 70% of customer power outages originate from distribution network faults, and the frequency and duration of these outages are important reliability indicators for measuring distribution network performance.

[0019] To improve power supply availability and shorten fault recovery time, an effective approach is to rationally configure sectionalizing switches. One commonly used method for planning sectionalizing switches is to evenly divide the number of users and loads, using this as a basis for laying out automated switches. This method is simple to operate and easy to implement, and can meet basic power grid operation requirements to a certain extent.

[0020] However, the currently used methods have the following technical problems: different users have the same electricity demand and load, and in the event of a distribution network fault, independent power supply areas (i.e., islands) are formed through distributed generation (DG), requiring DG to supply power to different users, thus causing the power supply structure of the distribution network to change constantly. Existing methods are difficult to adapt to users with different needs and the daytime variable distribution network structure, resulting in a large deviation between the configured sectionalizing switch layout and the actual situation, and requiring high configuration accuracy. This not only fails to meet current usage requirements but also increases the layout cost of sectionalizing switches.

[0021] To address the aforementioned issues, the following detailed embodiments will be used to illustrate and explain a method, apparatus, equipment, and medium for planning the layout of sectional switches in a power distribution network, as provided in this application.

[0022] To address the technical problems of large deviations between the planned and actual segmented switches in existing technologies, and low planning and matching accuracy, this paper refers to... Figure 1 The diagram shows a flowchart of a method for planning the layout of sectional switches in a power distribution network according to an embodiment of the present invention.

[0023] As an example, the method for planning the sectionalizing switch layout of the distribution network may include: S11. After obtaining multiple different switch layout schemes, the MIGOA algorithm is called to iterate the multiple switch layout schemes to obtain a local scheme set, wherein each of the switch layout schemes is a layout planning scheme of segmented switches in the distribution network.

[0024] In one embodiment, in order to ensure that the layout planning of segmented switches can meet actual needs and reduce deviations, multiple different switch placement schemes can be constructed in a conventional manner. By pre-constructing switch placement schemes, and then processing them by calling algorithms based on the prior switch placement schemes, the switch placement schemes that meet the actual load requirements can be selected, which can fit the actual needs and reduce deviations.

[0025] In one operating mode, after obtaining multiple different switch placement schemes, each scheme is a layout planning scheme for sectional switches in the distribution network, specifically a planning scheme for the placement locations of the sectional switches. The MIGOA algorithm can be called to iterate through multiple switch placement schemes, allowing them to be aggregated to form a local scheme set, where the local scheme set contains multiple switch placement schemes.

[0026] In one embodiment, when the MIGOA algorithm is invoked to iterate through multiple switch placement schemes, the number of iterations n is... p Reaching the preset maximum value N p The algorithm terminates when N is too large. p This could lead to a sharp increase in computational costs. Therefore, based on the above conditions, this invention introduces another termination condition N. pnb That is, in a continuous n pnb The iteration terminates when no better solution is found. When n... p Reaching its threshold N p , or when n pnb Reaching its threshold N pnb When that happens, the algorithm stops running.

[0027] As an example, the step of iterating through the multiple switch placement schemes using the MIGOA algorithm after obtaining multiple different switch placement schemes to obtain a local scheme set may include the following sub-steps: S111. Obtain multiple different switch placement schemes, and divide the multiple switch placement schemes into N preprocessed subgroups, wherein each subgroup contains several individuals, and each individual is a switch placement scheme.

[0028] To fully utilize computational resources, the algorithm employs a parallel GOA strategy. In this algorithm, the global population is divided into N subpopulations, each containing μ individuals, and they evolve independently. All individuals are generated randomly in the initial generation and represented using binary encoding.

[0029] It should be noted that the subpopulations are the solution space representation within the algorithm, which is the set of individuals (solutions), corresponding to the scheme set. Each individual represents a switch placement scheme, usually encoded in binary string, the length of which is the number of all possible installation positions of the switch in the scheme. A bit of 0 indicates that the switch is installed at the corresponding candidate position, and a bit of 1 indicates that it is not installed.

[0030] In an optional embodiment, obtaining multiple different switch placement schemes and dividing the multiple switch placement schemes into N preprocessed subpopulations may include the following sub-steps: S1111. Obtain multiple different switch placement schemes and divide the multiple switch placement schemes into N initial populations, wherein each initial population includes multiple individuals.

[0031] S1112. Preprocess each individual in the initial population to obtain a subpopulation, wherein the preprocessing is the process of removing similar individuals from each initial population.

[0032] There may be identical or similar switch placement schemes in the prior construction. When dividing, identical or similar switch placement schemes can be divided into a population.

[0033] Alternatively, several switch placement schemes can be divided into a population based on their quantity. For example, if there are 50 switch placement schemes in total, the 1st to 10th switch placement schemes can be in one population, the 11th to 20th switch placement schemes can be in another population, and so on.

[0034] Alternatively, these individuals can be evenly or randomly divided into N populations, each with a size of μ. Each population evolves independently until a termination condition (N) is met. p or N pnb ).

[0035] Using the above partitioning method, N initial populations can be obtained, where N is a positive integer greater than 1. Each initial population consists of one or more individuals, and each individual corresponds to a switch placement scheme.

[0036] Since the previously constructed switch placement schemes may have the same or similar schemes, repeating the same scheme for iteration will reduce processing efficiency. Preprocessing can be performed on each initial population to remove similar individuals within each initial population, thereby reducing the number of individuals in each initial population.

[0037] In an optional embodiment, the preprocessing of individuals within each initial population to obtain a subpopulation may include the following sub-steps: S11121. Randomly select at least one individual from each of the initial populations as a reference.

[0038] S11122. Calculate the Hamming distance between the reference body and each non-reference body in the initial population to obtain multiple individual distance values.

[0039] S11123. Identify and remove individuals in the initial population whose body distance values ​​satisfy the preset distance value to obtain a subpopulation.

[0040] In one embodiment, to ensure that individuals in the population are distributed as evenly as possible within the solution space, a crowding factor CF is set. A subset of individuals are randomly selected as reference individuals in each initial population. Then, within that initial population, the Hamming distance between each reference individual and every non-reference individual in the same initial population is calculated, resulting in multiple individual distance values. The similarity between individuals is determined using a Hamming distance threshold dm0.

[0041] Next, individuals within the initial population whose body distance values ​​meet the preset distance values ​​can be identified, and these individuals are then removed to obtain a subpopulation. Individuals that are highly similar to non-reference bodies are discarded or respawned to enhance the population's diversity and heterogeneity.

[0042] S112. Calculate the fitness of each individual in each subpopulation, and migrate individuals within multiple subpopulations based on the fitness until the individuals in each subpopulation converge, thereby obtaining a set of local solutions.

[0043] In one embodiment, the fitness of each individual in each subpopulation can be calculated, whereby the fitness of each individual's proposed solution is related to the planning objective. If the fitness is high, it indicates that the individual is well-suited to the planning objective of the segmented switch; conversely, if the fitness is low, it indicates that the individual is not well-suited to the planning objective of the segmented switch.

[0044] Individuals within each subpopulation can be migrated based on fitness until the population converges. For example, if the first individual in the first subpopulation has a high fitness, it can be migrated to a high-fitness subpopulation, `ExcellentPool`. Then, the fitness of the second individual in the first subpopulation can be calculated; if the fitness of the second individual is low, migration is not necessary. This process is repeated until all individuals in the first subpopulation have completed their fitness calculations. Then, the fitness of the first individual in the second subpopulation is calculated. If the fitness of the first individual in the second subpopulation is high, it is migrated to `ExcellentPool`. This process is repeated to adjust the individuals in each subpopulation until convergence. Individuals suitable for the planning objective can be selected from each subpopulation, resulting in a local solution set. The local solution set can also include N subpopulations and a constructed subpopulation `ExcellentPool`, where the number of individuals in each subpopulation may vary.

[0045] In an optional embodiment, calculating the fitness of each individual in each of the subpopulations, and migrating individuals within multiple subpopulations based on the fitness until the individuals in each subpopulation converge to obtain a set of local solutions, may include the following sub-steps: S1121. Calculate the fitness of each individual based on the shared function of each individual, and use the fitness to calculate the mutation probability and crossover probability of each individual.

[0046] S1122. Using the roulette wheel method, randomly select the corresponding individuals from each of the subpopulations as individuals to be processed.

[0047] S1123. Based on the mutation probability and crossover probability of the individuals to be processed, migrate the individuals to be processed in each subpopulation until the individuals in each subpopulation converge, thereby obtaining a local scheme set.

[0048] The fitness of each individual is calculated by a pre-defined objective function, since the objective function... F The fitness value needs to be minimized, and its initial fitness value is determined accordingly. To maintain diversity, a fitness sharing mechanism is introduced, and the fitness can be calculated as follows: ; In the above formula, C switch Defined as switching cost; , It is the weight of the cost item, N f N is the total number of feeders, T is the predicted lifespan of the sectionalizing switch, and N is the total number of feeders. q For feeder f The total number of segments q, N LPγ represents the total number of load points; γ is the load growth rate, calculated by multiplying the average load of the load points by γ. ; The interruption cost at load point j in year t due to the temporary shutdown of feeder section q in section f; The expected interruption cost for the user's system; CF1 and CF2 correspond to two items of the optimization objective, with the optimization objective as the fitness.

[0049] In one embodiment, the present invention uses a clustering mechanism based on a shared function to avoid the population getting trapped in local convergence. The shared function can be expressed as follows: ; in, Represents an individual and Shared functions, and Represent their Hamming distance and primary fitness interval, respectively; shared radius and Used to limit the maximum range of similarity.

[0050] Meanwhile, the fitness function was adjusted as follows. This adjustment weakens the fitness of individuals due to excessive similarity, thereby suppressing clustering. Specifically, the adjusted fitness can be calculated as follows: ; in, For the original fitness Its independent variable is x k M represents the size of the subpopulation.

[0051] After calculating the fitness of each individual, the corresponding mutation probability and crossover probability can also be calculated based on the fitness of the individual.

[0052] Where the crossover probability cp is in the interval Within this range, the interval values ​​of the crossover probability are shown in the following formula: ; Where cp represents the adaptive crossover probability; cp1 and cp2 are its set maximum and minimum values, respectively. F max and F avg These represent the maximum and average fitness of the population, respectively. For individuals that have undergone variation, F refers only to their primary fitness.

[0053] The mutation probability amp is in the interval Within this range, the interval values ​​of the mutation probability are shown in the following formula: ; Where amp represents the adaptive mutation probability; and These are the maximum and minimum values ​​set for them, respectively.

[0054] To ensure that every individual in the subpopulation has a probability of being selected, the classic roulette wheel method can be used for individual selection, such that every individual member in the subpopulation has a probability of being selected, and the probability of being selected can be expressed as follows: .

[0055] In the process of handling crossover and mutation, to avoid the limitations of fixed crossover and mutation probabilities, the core idea is: the crossover probability cp is within the interval... The mutation probability amp dynamically adjusts with individual fitness; the mutation probability amp is within the range Internal changes.

[0056] In the crossover operation, if the two individuals F A larger value indicates a higher crossover probability, allowing two individuals to interact within their respective populations. The crossover and mutation operations employ constant crossover and bit-flipping techniques, respectively. Individuals with lower fitness are more easily altered through crossover and mutation operators, while those with higher fitness are less susceptible to change. This technique enhances global search performance and improves GOA convergence efficiency.

[0057] After determining the probabilities of crossover and mutation, individual migration can be performed. MIGOA is one of the better-performing variants of GOA, introducing a migration operator in Multi-Island GOA (MIGOA) to maintain the diversity of subpopulations. Its basic mechanism is: dividing the global population into several discrete islands, and migrating the population every p... m At that time, according to the migration probability m pp A number of individuals are selected from some islands and migrated to other islands. Ultimately, each subpopulation may converge to a different local optimum, which is usually contained within the global optimum.

[0058] In each round of MIGOA iteration, the fitness of each individual can be calculated, and the top 10% of individuals (solutions) with the best fitness are selected to update the subpopulation ExcellentPool as the local optimization population. After completing the above steps of MIGOA iteration calculation, the resulting set of local solutions is the set of local optimal solutions for the segmented switch optimal configuration scheme, where the switch configuration scheme is the segmented switch placement location information.

[0059] S12. Call the SDT algorithm to search and iterate the solutions in the local solution set to obtain the convergent solution set.

[0060] In one embodiment, the local solution set includes several populations. Some populations may have solutions that are compatible with the planning, while others may not be compatible with the planning. The SDT algorithm can be called to search and iterate the solutions in the local solution set. Then, several compatible individuals can be searched from individuals in different populations and these individuals can be combined into a set to obtain the convergent solution set.

[0061] In an optional embodiment, the step of invoking the SDT algorithm to search and iterate the solutions in the local solution set to obtain a convergent solution set may include the following sub-steps: S121. Select one of the multiple subpopulations in the local scheme set as the target population, wherein the target population is the subpopulation in the local scheme set whose fitness satisfies a threshold.

[0062] S122. Calculate the gradient along the target population using the differential algorithm and determine the preset step size, wherein the preset step size is the search step size calculated based on the adaptive binary search process.

[0063] S123. Starting from the target population, search for the optimal population among multiple processing populations in the local scheme set according to the gradient and the preset step size, until the difference between the optimal population and the target population satisfies the preset convergence threshold, and use the individuals of the optimal population as the convergence scheme set.

[0064] In one embodiment, the SDT algorithm starts from an initial feasible solution x (0) Starting from the current solution x, during the iteration process... (j) Using the initial point as the base, a local search is performed along the gradient direction of the objective function F(x). As a local search method, the performance of SDT largely depends on the quality of the initial solution. A poor initial solution can easily lead to an invalid search or get stuck in local convergence.

[0065] Therefore, this invention can select one subpopulation from multiple subpopulations of the local solution set as the target population. This target population is the subpopulation in the local solution set whose fitness satisfies a threshold. As explained above, the target population can be the updated subpopulation `ExcellentPool`. This invention uses the updated subpopulation `ExcellentPool` as the starting point of the SDT to avoid invalid computation and improve the overall optimization effect. At any stage, it starts from the solution x of the final stage. (j) and along the gradient Search in the middle.

[0066] Therefore, a differential algorithm can be used to calculate the gradient along the target population. In one embodiment, the gradient calculation can be expressed as follows: ; ; In the above formula, t j This refers to the step size to be calculated, which is t using an adaptive bisection process. j ; It is obtained using the difference algorithm; e n It is a unit vector; It is the small perturbation quantity in each direction of the independent variable, also known as the difference step size.

[0067] When the step size t is determined j Then, starting with the target population, the optimal population is searched among multiple treatment populations in the local scheme set according to the gradient and the preset step size. The search can continue from a new point, and the search can be expressed as follows: ; After a series of repetitions, it is possible to Convergence is achieved under the following conditions, where This represents a defined step size threshold. SDT is a local search process whose effectiveness largely depends on the first solution, and its purpose is to avoid useless search costs and local convergence.

[0068] It should be noted that in each round of SDT iteration, the gradient descent direction of each individual in ExcellentPool is calculated using the above formula to update ExcellentPool.

[0069] In steps S121-123 above, each individual in ExcellentPool (i.e., the local optimal solution) is used as an initial feasible solution. It is iterated along the gradient calculation formula and the search calculation formula, and its corresponding optimal solution is locally searched along the gradient descent direction. ExcellentPool is then updated with the optimal solution to obtain a set of convergent solutions.

[0070] S13. Calculate the total cost of each scheme in the convergence scheme set based on the objective function of the preset target model to obtain multiple scheme cost values. The objective function of the preset target model is a function constructed from the downtime of load points inside and outside the island and the cost of sectionalizing the switch.

[0071] After obtaining the convergent solution set, which includes multiple individuals, each representing a switch layout scheme, the total cost for each scheme in the convergent solution set can be calculated based on the objective function of the preset target model, resulting in multiple scheme cost values.

[0072] The preset target model can be a segmented switch placement planning model. The objective function of the segmented switch placement planning model is to minimize the switching cost, the interruption cost of temporary line outages, and the user's expected system interruption cost. In one embodiment, the objective function can be expressed as follows: ; in, ; ; ; ; In the above formula, C switch Defined as switch cost, including capital investment cost C s Installation cost C I and maintenance costs C t MS C switch As shown in the above formula; N f N is the total number of feeders, T is the predicted lifespan of the sectionalizing switch, and N is the total number of feeders. q For feeder f The total number of segments q, N LP This represents the total number of load points; additionally, the load growth rate γ is considered, multiplying the average load of each load point by... ; x f,q The 0-1 variable is installed for the sectionalizing switch; if the switch is installed on the feeder... f The segment q on, then x f,q =0, otherwise x f,q =1; C represents the interruption cost at load point j in year t due to the temporary shutdown of feeder section q in section f. temp Interruption cost for temporary faults (RMB / kW); L jtf Let f be the average load of feeder f at load point j in year t. For feeder f Section q leads to load point j Temporary failure rate; The user's expected system interruption cost; The failure rate of the component in segment q; The cost of load point j being out of service due to the outage of feeder section f (q) is expressed in yuan / kW. For the duration of the shutdown, It can be obtained from the user's loss function; , It is the weight of the cost item; It is the product of the user loss function and the annual average load at the feeder load point; It's the cost of switch maintenance.

[0073] In one embodiment, constraints can be set for the objective function. One of the constraints in the switch optimization configuration problem is the number of available switches to be installed. Sectional switches are expensive to install, so increasing the number of switches installed in the distribution network will significantly increase switch costs. However, while improving reliability, it is necessary to optimize the switch locations based on budget constraints. The number of available switches to be installed can be expressed as follows: ; In the above formula, N as This represents the total number of available switches.

[0074] Islanding becomes possible when distributed generation (DG) does not fail and its generating capacity is higher than or at least equal to the load demand within the island. This also requires meeting the voltage limits of the load nodes and the capacity limits of the feeders (or segments).

[0075] The probability of an island being isolated is given by the following equation: ; In the above formula, P GDG P represents the cumulative probability that the generating power of a DG unit is equal to or greater than a certain value. GLP This represents the probability that the load point has the above values; P f It is the forced outage rate of DG units.

[0076] Assuming that loads within the island will continuously receive power during islanding, and considering that loads may be located within the islanded area with a certain probability, the failure rate and associated downtime within the island can be calculated as follows: ; ; In the above formula, and This refers to the failure rate and downtime within the isolated area. and The failure rate and downtime of units without DG (Dual Generator).

[0077] also, and These represent the failure rate and outage time at the load point when the DG unit is present. By substituting the formulas for failure rate and related outage time into the formula: In this context, the proposed MILP, as the objective function, can minimize the solution objective.

[0078] Assuming distributed generation (DG) is located at the end of each feeder at the load point, if a switch is installed and operational between line k and the fault point, the fault can be cleared. Under this assumption, the outage time at the load point equals the switching time, and the possible interruption time at the load point is relatively short. If there is no switch between line k and the faulty line, the outage time at the load point equals the repair time of the faulty line segment.

[0079] The calculation of the downtime of feeder f at load point j can be derived from the following equation: ; In the above formula, l k N is the length of line segment k. LPf It is the total number of load points in the feeder; r L With r SW These represent the average time for line fault repair and the switching time for isolating the faulty line, respectively. The failure rate is the failure rate per unit length of segment k, including line failure rate, switch failure rate, and user access point failure rate. l j+1 It is a line connected to load points j and j+1.

[0080] For load points not located within the island formed by the DG, the power outage time can be calculated as follows: ; In the above formula, x f,cs For the 0-1 variable of the tie switch, if the feeder f If there is a handover switch at the end, then x f,cs =0; N LPf It represents the total number of load points in the feeder.

[0081] Based on the probability of DG islanding operation, this invention provides a detailed method for calculating the outage time of load points inside and outside the island of the power distribution system. Since the outage time of users is very important for the assessment of reliability costs, the outage time of the above-mentioned load points and the outage time of the load points can provide an accurate assessment of reliability indicators.

[0082] If a distributed generation (DG) is added at any point on the feeder, and a fault occurs on a segment before the DG, the fault can be isolated by installing a switch between the faulty line and the DG. The DG will then be able to continuously supply power to the load points within the isolated system. The downtime of the load point is equal to the switching time. Otherwise, the load clearing time is equal to the repair time of the faulty line segment. When a fault occurs on a segment of the line between the DG and the load point, even with a switch installed, the downtime of the load point is equal to the repair time of the faulty line. When a fault occurs on a segment of the line after the load point, the fault can be cleared by installing and operating a switch between the line and the faulty line. In this case, the downtime of the load point is equal to the switching time. Otherwise, the downtime of the load point is equal to the repair time of the faulty line segment.

[0083] This invention uses the user's expected system interruption cost. E COST Interruption costs caused by temporary failures Constructing an objective function can link reliability with customer power outage costs, taking into account the impact of interruption duration, equipment failure rate, load changes, and user loss functions. The planning cost of sectionalizing switches involves a variety of considerations regarding installation and fault conditions. S14. Select a target cost value from multiple scheme cost values, determine the scheme corresponding to the target cost value as the target planning scheme, and perform layout processing of the sectional switches of the distribution network based on the target planning scheme.

[0084] In one embodiment, a target cost value that meets the requirements can be selected from multiple alternative cost values. The alternative corresponding to the target cost value is determined as the target planning scheme. The target planning scheme may include the location (and quantity) of sectionalizing switches. The layout and adjustment of sectionalizing switches in the distribution network can be performed based on the target planning scheme.

[0085] In one embodiment, the step of filtering a target cost value from multiple scheme cost values, determining the scheme corresponding to the target cost value as the target planning scheme, and performing layout processing of the sectional switches of the distribution network based on the target planning scheme may include the following sub-steps: S141. Select the scheme cost value with the smallest value from the multiple scheme cost values ​​as the target cost value, and determine the scheme corresponding to the target cost value as the target planning scheme.

[0086] S142. Determine the location information of each segment switch within the target planning scheme, and perform layout processing of the segment switches of the distribution network based on the location information.

[0087] The minimum cost value among multiple alternative cost values ​​is selected as the target cost value. The alternative with the minimum cost value is chosen as the final solution, and the alternative corresponding to the target cost value is determined as the target planning solution.

[0088] Then, determine the location information of each segment switch within the target planning scheme, and perform layout processing of the segment switches of the distribution network based on the location information.

[0089] It should be noted that for distribution networks that already have sectionalizing switches installed but still require further rational planning of their locations, the 0-1 variables for sectionalizing switch installation... x f,q The value is fixed at 0 for line sections with installed switches, meaning that existing switches are retained and the number and location of new switches are optimized only for candidate locations to avoid unnecessary removal, which is in line with engineering practice and saves costs. For distribution networks that have not yet been planned to install sectional switches, sectional switches will be installed later according to the optimal switch configuration scheme.

[0090] For distribution networks that have not yet been planned to install sectionalizing switches, the locations of the sectionalizing switches will be installed later according to the switch configuration plan.

[0091] Reference Figure 2 The diagram shows an operation flowchart of a segmented switch layout planning method for a power distribution network according to an embodiment of the present invention.

[0092] Specifically, the operation of the sectional switch layout planning method for a distribution network may include the following steps: Step 1: Data Processing: Distribution network topology, load level, DG data; reliability indicators (failure rate, repair time); weighting factors. , Algorithm parameters include the number of islands N, the population size of each island μ, and the maximum number of iterations N. p Maximum number of iterations N without improvement pnb wait.

[0093] The second step is to model the objective function using a segmented switch placement model.

[0094] The third step involves creating a new population through a segmented layout scheme, dividing the global population into N subpopulations, each containing μ individuals, which evolve independently.

[0095] Fourth step, iteration number n=1, n pnb =1, initialize the optimal solution set. Where n pnb This refers to the number of iterations that result in no improvement during the iteration process, and its initial value is set to 1.

[0096] The fifth step involves pre-selection, which utilizes the crowding mechanism from NGOA to ensure that individuals in the population are distributed as evenly as possible within the solution space.

[0097] The sixth step involves calculating the fitness of each individual in the population using the objective function of the segmented switch placement model, the fitness sharing mechanism, the sharing function, and the fitness function calculation formula. The top 10% of individuals with the smallest objective function are then used as the optimal solutions to update the ExcellentPool set.

[0098] The seventh step, the selection phase, uses the classic roulette wheel method for individual selection, ensuring that every member in the subpopulation has a probability of being selected.

[0099] The eighth step, the crossover and mutation stage, employs constant crossover and bit reversal techniques, and introduces formulas for calculating the intervals of crossover and mutation probabilities to enhance global search characteristics and GOA convergence efficiency.

[0100] Step 9, the migration phase, introduces a migration operator in multi-island GOA to maintain the diversity of subpopulations, so that each subpopulation converges to a different local optimum.

[0101] Step 10: Determine if the optimal solution has been improved. If the improvement is complete, determine if the termination condition (N) has been met. p or N pnb ).

[0102] Step 11, if the termination condition (N) is met. p or N pnb The SDT algorithm uses each individual in the ExcellentPool as the current solution and performs a local search along the gradient direction of the objective function F(x).

[0103] Step 12: If the solution is improved, update the individuals in ExcellentPool.

[0104] Step 13: Return the optimal segmented switch configuration and provide the total cost.

[0105] In this embodiment, the present invention provides a method for segmented switch layout planning in a distribution network. Its advantages are as follows: After obtaining multiple different switch placement schemes, the present invention uses the MIGOA algorithm to iterate through these schemes to obtain a local scheme set; then, the SDT algorithm is used to search and iterate through the schemes in the local scheme set to obtain a converged scheme set; based on the objective function of a preset target model, the total cost of each scheme in the converged scheme set is calculated to obtain multiple scheme cost values; a target planning scheme is selected from the multiple scheme cost values, and the segmented switches in the distribution network are then laid out based on the target planning scheme. The present invention can obtain multiple different switch placement schemes, and through the MIGOA and SDT algorithms, several switch placement schemes that meet the requirements of the distribution network can be selected. The cost of each scheme is calculated using a cost-based objective function, and the scheme is then selected based on the cost. This not only allows for the selection of schemes that meet the requirements of the distribution network, reducing the deviation between the segmented switch layout and the actual situation, and improving configuration accuracy, but also reduces the layout cost of the segmented switches.

[0106] This invention also provides a sectional switch layout planning device for a power distribution network, see [link to relevant documentation]. Figure 3 The diagram shows a schematic diagram of a sectional switch layout planning device for a power distribution network according to an embodiment of the present invention.

[0107] As an example, the distribution network segmented switch layout planning device may include: The iteration module 201 is used to call the MIGOA algorithm to iterate the multiple switch layout schemes after obtaining multiple different switch layout schemes, and obtain a local scheme set, wherein each switch layout scheme is a layout planning scheme of segmented switches in the distribution network. The search iteration module 202 is used to call the SDT algorithm to search and iterate the solutions in the local solution set to obtain a converged solution set. The calculation module 203 is used to calculate the total cost of each scheme in the convergence scheme set based on the objective function of the preset target model to obtain multiple scheme cost values. The objective function of the preset target model is a function constructed from the downtime of load points inside and outside the island and the cost of sectionalizing the switch. The layout planning module 204 is used to filter target cost values ​​from multiple scheme cost values, determine the scheme corresponding to the target cost value as the target planning scheme, and perform layout processing on the sectional switches of the distribution network based on the target planning scheme.

[0108] Optionally, after obtaining multiple different switch placement schemes, the MIGOA algorithm is invoked to iterate through the multiple switch placement schemes to obtain a local scheme set, including: Multiple different switch placement schemes are obtained, and the multiple switch placement schemes are divided into N preprocessed subgroups, wherein each subgroup contains several individuals, and each individual is a switch placement scheme. Calculate the fitness of each individual in each of the subpopulations, and migrate individuals within the multiple subpopulations based on the fitness until the individuals in each subpopulation converge, thus obtaining a set of local solutions.

[0109] Optionally, the step of obtaining multiple different switch placement schemes and dividing the multiple switch placement schemes into N preprocessed subpopulations includes: Multiple different switch placement schemes are obtained, and the multiple switch placement schemes are divided into N initial populations, wherein each initial population includes multiple individuals; Each individual in the initial population is preprocessed to obtain a subpopulation, wherein the preprocessing is the process of removing similar individuals from each of the initial populations.

[0110] Optionally, the preprocessing of individuals within each initial population to obtain a subpopulation includes: At least one individual is randomly selected from each of the initial populations as a reference. Calculate the Hamming distance between the reference body and each non-reference body in the initial population to obtain multiple individual distance values; Individuals whose body distance values ​​satisfy a preset distance value within the initial population are identified and removed to obtain a subpopulation.

[0111] Optionally, the step of calculating the fitness of each individual in each of the subpopulations, and performing mutual migration of individuals within multiple subpopulations based on the fitness, until the individuals in each subpopulation converge to obtain a set of local solutions, includes: The fitness of each individual is calculated based on the shared function of each individual, and the mutation probability and crossover probability of each individual are calculated using the fitness. The roulette wheel method is used to randomly select corresponding individuals from each subpopulation as individuals to be processed. Based on the mutation probability and crossover probability of the individuals to be processed, the individuals to be processed in each subpopulation are migrated until the individuals in each subpopulation converge, thus obtaining a local scheme set.

[0112] Optionally, the step of invoking the SDT algorithm to search and iterate through the local solution set to obtain a convergent solution set includes: Select one subpopulation from the multiple subpopulations in the local solution set as the target population, wherein the target population is the subpopulation in the local solution set whose fitness satisfies a threshold. The gradient is calculated based on the target population using a differential algorithm, and a preset step size is determined, wherein the preset step size is a search step size calculated based on an adaptive binary search process; Starting with the target population, the optimal population is searched among multiple processing populations in the local scheme set according to the gradient and the preset step size until the difference between the optimal population and the target population meets the preset convergence threshold, and the individuals of the optimal population are used as the convergence scheme set.

[0113] Optionally, the step of filtering a target cost value from multiple scheme cost values, determining the scheme corresponding to the target cost value as the target planning scheme, and performing layout processing of the sectional switches of the distribution network based on the target planning scheme includes: The minimum cost value among the multiple proposed cost values ​​is selected as the target cost value, and the proposed solution corresponding to the target cost value is determined as the target planning solution. Determine the location information of each segment switch within the target planning scheme, and perform layout processing of the segment switches of the distribution network based on the location information.

[0114] Those skilled in the art will understand that, for ease of description and brevity, the specific working process of the device described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0115] Furthermore, this application also provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the segmented switch layout planning method for the power distribution network as described in the above embodiments.

[0116] Furthermore, embodiments of this application also provide a computer-readable storage medium storing a computer-executable program for causing a computer to execute the distribution network segmented switch layout planning method as described in the above embodiments.

[0117] In the description of the embodiments of the present invention, it should be noted that the terms "above," "below," etc., indicating the orientation or positional relationship are based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing the embodiments of the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. When an element such as a layer, region, or substrate is referred to as being "above" or "on top of" another element, it may be directly on the other element, or there may be an intermediate element. Conversely, when an element is referred to as being "directly on" or "above" another element, there is no intermediate element. It should also be understood that when an element is referred to as being "below" or "under" another element, it may be directly below or under the other element, or there may be an intermediate element. Conversely, when an element is referred to as being "directly below" or "under" another element, there is no intermediate element. Unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0118] Those skilled in the art will understand that embodiments of this application may also include computer program products. Therefore, this application may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0119] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), devices, and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0120] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0121] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0122] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for planning the layout of sectional switches in a power distribution network, characterized in that, The method includes: After obtaining multiple different switch placement schemes, the MIGOA algorithm is called to iterate the multiple switch placement schemes to obtain a local scheme set, wherein each of the switch placement schemes is a layout planning scheme for segmented switches in the distribution network. The SDT algorithm is invoked to search and iterate over the solutions in the local solution set to obtain a convergent solution set; The total cost of each scheme in the convergence scheme set is calculated based on the objective function of the preset target model to obtain multiple scheme cost values. The objective function of the preset target model is a function constructed from the downtime of load points inside and outside the island and the cost of sectionalizing the switch. Select a target cost value from multiple proposed cost values, determine the proposed solution corresponding to the target cost value as the target planning solution, and perform layout processing on the sectional switches of the distribution network based on the target planning solution.

2. The method for planning the layout of sectional switches in a power distribution network according to claim 1, characterized in that, After obtaining multiple different switch placement schemes, the MIGOA algorithm is called to iterate through the multiple switch placement schemes to obtain a local scheme set, including: Multiple different switch placement schemes are obtained, and the multiple switch placement schemes are divided into N preprocessed subgroups, wherein each subgroup contains several individuals, and each individual is a switch placement scheme. Calculate the fitness of each individual in each of the subpopulations, and migrate individuals within the multiple subpopulations based on the fitness until the individuals in each subpopulation converge, thus obtaining a set of local solutions.

3. The method for planning the layout of sectional switches in a power distribution network according to claim 2, characterized in that, The process involves acquiring multiple different switch placement schemes and dividing these schemes into N preprocessed subpopulations, including: Multiple different switch placement schemes are obtained, and the multiple switch placement schemes are divided into N initial populations, wherein each initial population includes multiple individuals; Each individual in the initial population is preprocessed to obtain a subpopulation, wherein the preprocessing is the process of removing similar individuals from each of the initial populations.

4. The method for planning the layout of sectional switches in a power distribution network according to claim 3, characterized in that, The step of preprocessing individuals within each initial population to obtain a subpopulation includes: At least one individual is randomly selected from each of the initial populations as a reference. Calculate the Hamming distance between the reference body and each non-reference body in the initial population to obtain multiple individual distance values; Individuals whose body distance values ​​satisfy a preset distance value within the initial population are identified and removed to obtain a subpopulation.

5. The method for planning the layout of sectional switches in a power distribution network according to claim 2, characterized in that, The process of calculating the fitness of each individual in each subpopulation, and performing mutual migration of individuals within multiple subpopulations based on the fitness, until the individuals in each subpopulation converge, to obtain a set of local solutions, includes: The fitness of each individual is calculated based on the shared function of each individual, and the mutation probability and crossover probability of each individual are calculated using the fitness. The roulette wheel method is used to randomly select corresponding individuals from each subpopulation as individuals to be processed. Based on the mutation probability and crossover probability of the individuals to be processed, the individuals to be processed in each subpopulation are migrated until the individuals in each subpopulation converge, thus obtaining a local scheme set.

6. The method for planning the layout of sectional switches in a power distribution network according to claim 2, characterized in that, The step of invoking the SDT algorithm to iterate through the local solution set to obtain a convergent solution set includes: Select one subpopulation from the multiple subpopulations in the local solution set as the target population, wherein the target population is the subpopulation in the local solution set whose fitness satisfies a threshold. The gradient is calculated based on the target population using a differential algorithm, and a preset step size is determined, wherein the preset step size is a search step size calculated based on an adaptive binary search process; Starting with the target population, the optimal population is searched among multiple processing populations in the local scheme set according to the gradient and the preset step size until the difference between the optimal population and the target population meets the preset convergence threshold, and the individuals of the optimal population are used as the convergence scheme set.

7. The method for planning the layout of sectional switches in a power distribution network according to any one of claims 1-6, characterized in that, The step of filtering a target cost value from multiple proposed cost values ​​and determining the proposed solution corresponding to the target cost value as the target planning solution, and then performing layout processing of the sectional switches of the distribution network based on the target planning solution, includes: The minimum cost value among the multiple proposed cost values ​​is selected as the target cost value, and the proposed solution corresponding to the target cost value is determined as the target planning solution. Determine the location information of each segment switch within the target planning scheme, and perform layout processing of the segment switches of the distribution network based on the location information.

8. A sectional switch layout planning device for a power distribution network, characterized in that, The device includes: The iteration module is used to call the MIGOA algorithm to iterate the multiple switch layout schemes after obtaining multiple different switch layout schemes, and obtain a local scheme set, wherein each switch layout scheme is a layout planning scheme of segmented switches in the distribution network. The search iteration module is used to call the SDT algorithm to search and iterate the solutions in the local solution set to obtain a converged solution set. The calculation module is used to calculate the total cost of each scheme in the convergence scheme set based on the objective function of the preset target model to obtain multiple scheme cost values. The objective function of the preset target model is a function constructed from the downtime of load points inside and outside the island and the cost of sectionalizing the switch. The layout planning module is used to filter target cost values ​​from multiple scheme cost values, determine the scheme corresponding to the target cost value as the target planning scheme, and perform layout processing on the sectional switches of the distribution network based on the target planning scheme.

9. An electronic device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, when the processor executes the computer program, it implements the method for segmented switch layout planning of a power distribution network as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer-executable program for causing a computer to perform the segmented switch layout planning method for a power distribution network as described in any one of claims 1-7.

Images