County power distribution network grid planning method based on digital resource fusion technology, and system

Abstract

Disclosed in the present invention are a county power distribution network grid planning method based on a digital resource fusion technology, and a system. The method comprises: acquiring geographic information of a county region, and performing preliminary division of power supply grids; performing secondary division of the power supply grids by means of a K-means clustering algorithm; and by means of an adaptive multi-objective whale optimization algorithm, selecting an optimal planning strategy to perform tertiary division of the power supply grids. By studying the development and evolution mechanism of a county power distribution network across temporal, spatial, and resource dimensions, planning and operation methods for the county power distribution network are integrated for digital resource fusion, enabling grid division of the county power distribution network. Moreover, a multi-objective collaborative planning model of the power distribution network is constructed from the perspectives of economy and security, and an optimal planning solution is selected.

Description

County distribution network grid planning method and system based on resource digital fusion technology Technical Field

[0001] The present invention relates to the technical field of grid planning of power grids, and in particular to a method and system for grid planning of county distribution networks using digital resource fusion technology. Background Art

[0002] With the rapid advancement of smart grid strategies, renewable energy generation, such as solar and wind power, is increasingly replacing traditional thermal power generation. However, the integration of distributed resources presents numerous challenges and issues for distribution network planning. Grid-based planning is the future direction of distribution network planning. The grid-based approach aims to reduce the complexity and dimensionality of distribution network control by simplifying and managing the distribution network. This approach is a key way to reduce the complexity and dimensionality of distribution network control.

[0003] Traditional grid division methods only consider topography, without considering the impact of distributed resource access. This approach, building on traditional grid division, proposes a county-level distribution network grid planning strategy based on resource digital fusion technology, tailored to the varying topography, load conditions, and distributed resource development within different counties. This strategy provides scenario-based, differentiated, and standardized planning for the distribution network. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is that the existing county distribution network grid planning method has problems such as high cost and unreasonable division.

[0006] To solve the above technical problems, the present invention provides the following technical solution: a county distribution network grid planning method based on resource digital fusion technology, comprising:

[0007] Collect geographic information of county areas and conduct preliminary division of power supply grids;

[0008] The power grid is divided for the second time using the K-means clustering algorithm;

[0009] The adaptive multi-objective whale optimization algorithm is used to select the best planning strategy and perform the third division of the power grid.

[0010] As a preferred solution of the multi-objective county distribution network optimization method described in the present invention, the preliminary division includes the first division of the power supply grid on the power grid digital platform according to the geographical conditions of mountains, waters, transportation, and different functional areas within the county.

[0011] As a preferred solution of the multi-objective county distribution network grid optimization method described in the present invention, wherein: the K-means clustering algorithm includes determining the number of clusters k in the county area based on the preliminary grid division results, and using the grid geometric center as the center point of each cluster;

[0012] Determine the classification conditions, and use the power supply distance between the substation and the load point and the distributed resources as the classification conditions;

[0013] Based on the measurement of power supply range, each node is compared with the nearest cluster center and assigned to the cluster with the closest distance;

[0014] Recalculate the data points in each cluster, that is, the mean of the node objects in each cluster, and calculate the clustering results of all clusters at the same time;

[0015] If the result value changes, repeat the steps of determining the classification conditions, otherwise the algorithm ends and outputs the second mesh division result.

[0016] As a preferred solution of the multi-objective county distribution network grid optimization method of the present invention, the third division includes establishing a county distribution network grid multi-objective coordinated planning model, and achieving the third division by solving the multi-objective optimization model;

[0017] The county-level distribution network grid-based multi-objective coordinated planning model includes the following: total grid planning investment cost:

[0018] Among them, C eco represents the total investment cost of grid planning; ε=k1+k2, where k1 and k2 are the operation and maintenance cost coefficient and the depreciation investment recovery coefficient respectively; C ij,m represents the total cost of the m-th inter-station power supply grid trunk line, which is the comprehensive cost of the entire length of all trunk lines; C ih,n represents the comprehensive cost of the trunk line of the self-loop power supply unit in the nth non-inter-station power supply grid; C fs,n represents the comprehensive cost of the trunk line of the radial power supply unit in the nth non-inter-station power supply grid; represents the annual cost of power loss on the main line of the m-th inter-station power supply grid; represents the annual cost of power outage loss on the trunk line of the nth non-inter-station power supply grid; C in represents the investment and construction cost of distributed resources; C op represents the operation and maintenance cost of distributed resources; C pro represents the distributed resource benefit; C l represents the investment in digital communication between grids;

[0019] Assume that the interconnection line is built along the power line, and each node has a digital interconnection substation, and the interconnection substation is not far from the substation; set the digital interconnection investment to:

[0020] Where N represents the number of grids, d m represents the branch line length of grid m, p l Indicates the unit price of data transmission line, p c Indicates the unit price of the grid control unit;

[0021] The power outage loss describes the overall reliability index of the distribution network and is expressed as:

[0022] Among them, C los It is the reliability index of the distribution network, indicating the total power outage loss of the entire network; P lo,m,j represents the power outage probability of the jth node in the grid m; N los represents the number of power outages in the distribution network during the planning period; t represents the duration of a single power outage; N represents the number of load nodes in each grid; S m,j is the power supply status of load node j in grid m, 0 means power failure, 1 means normal; P m,j P (t) is the actual power consumed by load node j in grid m at time t; m,j (t) = P den,m,j (t)-P dis,m,j (t)

[0023] Among them, P den,m,j (t) represents the power demand of load node j in grid m at time t; P dis,m,j (t) represents the power generated by the distributed resources of load node j in grid m at time t;

[0024] Set the inter-grid power interaction objective function:

[0025] Among them, P ij represents the active power flow of branch i between grids, Q ij represents the reactive power flow of inter-grid branch i, k represents the number of inter-grid branches, P p 、P q represents the power interaction coefficient.

[0026] As a preferred solution of the multi-objective county distribution network optimization method of the present invention, wherein: the multi-objective optimization model solution includes, the mathematical model of the multi-objective optimization problem is expressed as: minF(x)=(C eco ,C los ,E m,n )

[0027] Introducing the Pareto dominance relationship, for m objective functions f i (x), i=1,2,...,m Given any two decision variables x1 and x2, if they satisfy the following equation, then the solution x1 is said to dominate x2;

[0028] The whale's hunting strategy is a process of continuously approaching the optimal solution to the optimization problem, where each whale can be regarded as a solution to the optimization problem, and the prey can be regarded as the optimal solution to be found. The optimization problem is solved by imitating the whale's hunting process.

[0029] Surrounding the prey: In the optimization problem, the individual with the best fitness function is regarded as the optimal individual in the population. The process of other individuals in the population changing their positions towards the optimal individual is expressed by the mathematical model as follows:

[0030] Where t represents the current number of iterations, Indicates the current position of the whale, Indicates the currently obtained optimal whale position, Indicates the whale's position after the updated position. represents the distance the whale moves, A and C are the correlation coefficients; A=2aR1-a, C=2R2, a=2(t max -t) / t max

[0031] Among them, R1 and R2 represent random floating point numbers on [0, 1], t max Indicates the maximum number of iterations set, and a indicates that the convergence factor will gradually decrease to 0 as the number of iterations increases;

[0032] Bubble net hunting involves spiral swimming: whales will swim in spirals towards their prey, which can be expressed mathematically as follows:

[0033] in, represents the distance between the whale and the optimal individual; b represents a constant used to define the spiral shape, usually 1; l is a random floating point number on (-1, 1).

[0034] As a preferred solution of the multi-objective county-level distribution network optimization method described in the present invention, the predation strategy also includes introducing an action probability P to determine whether the whale chooses to approach the prey by shrinking and encircling or swimming in a spiral. When the generated random number is less than P, the whale individual performs shrinking and encircling, and when it is not less than P, it performs spiral swimming. The mathematical model of whale bubble net hunting is:

[0035] Where p represents a random floating point number in the interval [0, 1]; P represents the action probability constant, which is usually set to 0.5;

[0036] Search hunting: Whales search for prey globally, and its mathematical model is expressed as:

[0037] in, represents the position of a randomly selected whale in the population;

[0038] When A≥1, the whale randomly selects individuals in the population to update their positions, enhancing the global search capability of the algorithm.

[0039] As a preferred solution of the multi-objective county distribution network optimization method described in the present invention, the mathematical model for calculating grid information is expressed as follows:

[0040] Where M represents the grid number of the individual, [·] represents the largest integer smaller than the number in the brackets, Grid represents the number of grids to be divided, t represents the number of iterations, and y i (t) represents the objective function value of individual i in the t-th generation, maxY(t) represents the maximum value of the corresponding objective function among all individuals in the t-th generation population, and minY(t) represents the minimum value of the corresponding objective function among all individuals in the t-th generation population;

[0041] Get the grid number tuples of all individuals, use the roulette wheel method to select the best individual, and randomly select one of the individuals as the best individual when the selected grid has multiple individuals.

[0042] A multi-objective county distribution network grid optimization system using the method of the present invention is characterized by:

[0043] The first division unit collects geographic information of the county area and performs preliminary division of the power supply grid;

[0044] The second division unit divides the power grid for the second time through the K-means clustering algorithm;

[0045] The third division unit selects the best planning strategy through the adaptive multi-objective whale optimization algorithm and divides the power supply grid for the third time.

[0046] A computer device comprises: a memory and a processor; the memory stores a computer program, wherein the processor implements the steps of any one of the methods of the present invention when executing the computer program.

[0047] A computer-readable storage medium stores a computer program thereon, wherein the computer program, when executed by a processor, implements the steps of any one of the methods of the present invention.

[0048] Beneficial effects of the present invention: The multi-objective county distribution network grid optimization method provided by the present invention performs a preliminary division of the power supply grid on the power grid digital platform according to the county's geographical conditions such as mountains, waters, transportation, and different functional areas. Taking into account the grid structure of the divided power supply grid, the power supply range of the substation, and the current development status of distributed resources, the power supply grid is divided for the second time according to the saturated load forecast results through the K-means clustering algorithm. With the goals of improving the absorption and utilization rate of distributed resources, reducing planning funds, and improving the reliability of the county distribution network, a multi-objective collaborative planning model for the distribution network is constructed, and the optimal planning strategy is selected through the adaptive multi-objective whale optimization algorithm to divide the power supply grid for the third time. Through the development and evolution mechanism of the county distribution network in time, space, and resource dimensions, the planning and operation methods of the county distribution network are combined to perform digital fusion of resources and divide the county distribution network into grids. A multi-objective collaborative planning model for the distribution network is constructed from the perspective of economy and safety to select the optimal planning scheme. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort. Among them:

[0050] FIG1 is an overall flow chart of a county-level distribution network grid planning method using a resource digital fusion technology according to a first embodiment of the present invention;

[0051] FIG2 is a Pareto optimal frontier diagram of the county distribution network grid planning method using the resource digital fusion technology provided by the first embodiment of the present invention;

[0052] FIG3 is a diagram showing the steps for implementing a multi-objective grid optimization algorithm for a county-level distribution network grid planning method using a resource digital fusion technology provided in a second embodiment of the present invention. DETAILED DESCRIPTION

[0053] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0054] Example 1

[0055] 1 , an embodiment of the present invention provides a county-level distribution network grid planning method using resource digital fusion technology, including:

[0056] S1: Collect geographic information of the county area and make a preliminary division of the power supply grid.

[0057] Furthermore, the preliminary division includes the first division of the power supply grid on the power grid digital platform according to the geographical conditions of mountains, waters, transportation, and different functional areas within the county.

[0058] The key point is that the collected GIS data and functional zone information are input into the system using the power grid digital platform. Within the platform, power supply demand and grid layout are simulated based on geographic information and functional zone characteristics. The impact of natural obstacles such as mountains and water on grid construction, as well as the convenience of transportation networks for grid maintenance, are considered. Based on the simulation results, a preliminary division of the power grid is performed. This step considers the grid's reliability, cost-effectiveness, and future scalability. The divided grid should effectively cover different functional zones, while also accounting for natural geographical constraints.

[0059] S2: The power grid is divided for the second time using the K-means clustering algorithm.

[0060] The K-means algorithm is a clustering algorithm based on an objective function, widely used in data mining and machine learning. This algorithm uses the method of finding the extreme value of a function to adjust the rules of the objective function. The objective function is distance optimization. The algorithm aims to minimize the evaluation function, based on the standard sum of squared errors, uses Euclidean distance as the similarity calculation, and starts classification with the initial distance center. The criterion function uses the sum of squared errors for clustering. When the number of cluster categories is known, the K-means algorithm can calculate and group a disorganized data set into K category clusters, which contain data samples of all categories.

[0061] The K-means algorithm is a clustering algorithm based on the number of clusters k. Its specific process is as follows: First, k cluster centers are randomly selected, and the similarity between the data point and the cluster center is calculated, and each data point is assigned to the nearest cluster center. Then, the center point of each cluster is recalculated, and the similarity between the data point and the cluster center is calculated again, and the data point is redistributed. This process is iterated until the clustering result is stable and no longer changes, and finally the cluster to which each data point belongs is obtained.

[0062] Here are the steps:

[0063] Step 1: Based on the preliminary grid division results, determine the number of clusters k in the county area, and use the grid geometric center (substation site) as the center point of each cluster;

[0064] Step 2: Determine the classification conditions, using the power supply distance between the substation and the load point and the distributed resource as the classification condition;

[0065] Step 3: Compare each node with the nearest cluster center based on the power supply range measurement method and assign it to the cluster with the closest distance;

[0066] Step 4: Recalculate the data points in each cluster, that is, the mean of the node objects in each cluster, and calculate the clustering results of all clusters at the same time;

[0067] Step 5: If the result value changes, repeat step 2; otherwise, the algorithm ends and the second meshing result is given.

[0068] Going further, the impact of distributed resource access on county distribution networks: analyze the impact of source-load operating characteristics on distribution networks from three dimensions: source-load category, source-load distribution location, and time scale, digitally integrate grid resources, rationally allocate distributed resources, and improve the absorption rate.

[0069] S3: Through the adaptive multi-objective whale optimization algorithm, the best planning strategy is selected to divide the power grid for the third time.

[0070] A grid-based multi-objective coordinated planning model for county-level distribution networks was established, and the third partitioning was achieved by solving the multi-objective optimization model. Substation locations and their power supply ranges, as well as candidate trunk channel layouts and load locations, were obtained on the grid digital platform. Under the conditions of satisfying the independent connectivity of each grid channel and the maximum allowable load transfer distance, the goal was to maximize the number of power supply partitions that could achieve load transfer while minimizing the total cost of grid investment. The corresponding multi-objective mixed integer nonlinear programming optimization model can be expressed as follows:

[0071] Where C eco represents the total investment cost of grid planning; ε=k1+k2, where k1 and k2 are the operation and maintenance cost coefficient and the depreciation investment recovery coefficient respectively; C ij,m represents the total cost of the m-th inter-station power supply grid trunk line, which is the comprehensive cost of the entire length of all trunk lines; C ih,n represents the comprehensive cost of the trunk line of the self-loop power supply unit in the nth non-inter-station power supply grid; C fs,n represents the comprehensive cost of the trunk line of the radial power supply unit in the nth non-inter-station power supply grid; represents the annual cost of power loss on the main line of the m-th inter-station power supply grid; represents the annual cost of power outage loss on the trunk line of the nth non-inter-station power supply grid; C in represents the investment and construction cost of distributed resources; C op represents the operation and maintenance cost of distributed resources; C pro represents the distributed resource benefit; C l Represents the investment in digital communication between grids.

[0072] The planning of digital interconnection is factored into the steady-state grid division. Reasonable digital interconnection facilities can improve the grid control effectiveness of the distribution network. Digital interconnection facilities consist of data transmission lines and grid control units. The following assumptions are made for digital interconnection facilities: interconnection lines are laid along power lines, digital interconnection substations exist at each node, and the interconnection substations are not far from the substations. The digital interconnection investment is set to:

[0073] Where: N is the number of grids; d m is the branch line length of grid m. l is the unit price of data transmission line; p c is the unit price of the grid control unit.

[0074] It should be noted that reliability includes: With the access of distributed resources, the overall reliability index of the distribution network can be obtained by accumulating the reliability index of each grid, and the power outage loss is described as:

[0075] Where: C los is the reliability index of the distribution network, which represents the total power outage loss of the entire network; P lo,m,j is the power outage probability of the jth node in grid m; N los is the number of power outages in the distribution network during the planning period; t is the duration of a single power outage; N is the number of load nodes in each grid; S m,j is the power supply status of load node j in grid m (0 means power failure, 1 means normal); P m,j (t) is the actual power consumed by load node j in grid m at time t.

[0076] Given the self-healing capability of each grid after the county distribution network is gridded, it can actively isolate the fault in the event of a fault and use distributed resources to achieve self-healing. The actual power consumption of each node during the fault recovery period is P m,j (t) can be further described as: P m,j (t) = P den,m,j (t)-P dis,m,j (t) (4)

[0077] Where: P den,m,j (t) is the power demand of load node j in grid m at time t; P dis,m,j(t) is the power generated by the distributed resources of load node j in grid m at time t.

[0078] Inter-grid power interaction includes: if the inter-grid power interaction is minimized, then the coordinated dispatch of the grid will be easier during the actual operation of the grid. Therefore, the inter-grid power interaction objective function is set as:

[0079] Where, P ij , Q ij are the active power flow and reactive power flow of inter-grid branch i respectively; k is the number of inter-grid branches; P p 、P q is the power interaction coefficient.

[0080] It is important to know that constraints include:

[0081] Distribution network flow constraints:

[0082] Where: P i , Q i are the active power and reactive power of node i respectively; P max,i , Q max,i are the maximum allowable values ​​of active power and reactive power of node i respectively; U i 、U j are the voltages at nodes i and j respectively; B ij , G ij are the conductance and susceptance between nodes i and j respectively; θ ij is the voltage phase angle between nodes i and j.

[0083] Node voltage constraints. U i,min ≤U i ≤U i,max (7)

[0084] Where: U i,min 、U i,max are the lower and upper limits of the voltage at node i, respectively.

[0085] Solving the multi-objective optimization model: The mathematical model of the multi-objective optimization problem is expressed as: minF(x)=(C eco, C los, E m,n ) (8)

[0086] The goal of multi-objective optimization is to find the optimal solution. However, in practical applications, these objective functions are often contradictory. Optimizing one objective function comes at the expense of other objective functions, making it difficult to achieve an ideal state of resource allocation. Furthermore, such an ideal state often has multiple solutions. Finding all such solutions within the solution space is the primary challenge facing multi-objective optimization algorithms.

[0087] To solve the multi-objective optimization problem, we must first introduce the concept of Pareto dominance relationship. For m objective functions f i (x),i=1,2,...,m Given any two decision variables x1 and x2, if they satisfy equation (9), then the solution x1 is said to dominate x2.

[0088] As shown in Figure 2, let the objective functions f1 and f2 be as small as possible, and A, B, C, D, E, and F are the six solutions in the objective space. Among them, B is better than E in both objective functions. According to Equation (2), B can be said to dominate E. Similarly, C can also be said to dominate F. As for A and D, since there are no solutions that dominate them in the solution space, the hypersurface where A, B, C, and D are located constitutes the Pareto front in the current solution space, as shown by the curve in the figure.

[0089] Therefore, if no other decision variable can dominate a decision variable, then this decision variable is called a non-dominated solution. Non-dominated sorting involves grouping all decision variables that meet this condition into a set, forming the Pareto front solution set, which is marked as the first layer. Then, all non-dominated solutions found are eliminated and the remaining individuals are sorted again. The resulting Pareto front solution set is marked as the second layer, and so on until all individuals in the population have been processed.

[0090] Furthermore, the whale optimization algorithm simulates the natural hunting process of whales: upon discovering prey, whales form a group, surround it, and swim toward it in a spiral motion, continuously emitting bubbles as they do so. This creates a cylindrical "bubble net" that tightly surrounds the prey until the whale closes in and swallows it whole. A whale's hunting strategy can be viewed as a process of continuously approaching the optimal solution to an optimization problem, where each whale can be considered a solution to the problem, and the prey can be considered the optimal solution to be found. Therefore, the whale's hunting process can be mimicked to solve the optimization problem.

[0091] Surrounding prey: Whales are social animals. Once an individual in the population spots prey, the others will move toward it to compete for it. In the optimization problem, the individual with the best fitness function is considered the optimal individual in the population. The process of other individuals in the population changing their positions toward the optimal individual is represented by a mathematical model:

[0092] Where: t is the current iteration number; is the current whale's position; is the best whale position currently obtained; The whale's position after the update; represents the distance the whale moves. A and C are correlation coefficients, as shown in formula (11): A=2aR1-a, C=2R2, a=2(t max -t) / t max (11)

[0093] Where: R1 and R2 are random floating point numbers on [0, 1]; t max is the maximum number of iterations set; a is the convergence factor, which gradually decreases to 0 as the number of iterations increases.

[0094] "Bubble Net" Hunting: Whales spit out "bubble nets" to surround their prey during the hunting process, so two mathematical models are designed to represent this hunting behavior. Contraction and Encirclement. Formula (3) is used to continuously contract and approach the prey. The difference from Formula (3) lies in the value range of A. Since whales only release "bubble nets" to trap prey when they are close to the prey, and A gradually decreases with the number of iterations, it is set here that when A < 1, the whale's updated position can be anywhere between it and the optimal individual.

[0095] Spiral swimming: Whales swim in spirals towards their prey, which can be represented by a mathematical model:

[0096] in: represents the distance between the whale and the optimal individual; b is a constant used to define the spiral shape, usually 1; l is a random floating point number on (-1, 1).

[0097] The two hunting behaviors described above occur during the predation process of different whales. To better simulate the asynchronous nature of these behaviors, an action probability P is introduced to determine whether the whale chooses to close in or spiral toward its prey. When the generated random number is less than P, the whale closes in and spirals out. When the generated random number is not less than P, the whale spirals out. Therefore, the mathematical model of whale "bubble net" hunting is:

[0098] Where: p is a random floating point number in the interval [0, 1]; P is the action probability constant, usually set to 0.5.

[0099] Search hunting: Whales can also search for prey globally. The mathematical model is expressed as:

[0100] in: The whale position is randomly selected in the population. When A≥1, the whale will randomly select individuals in the population to update their positions, enhancing the global search capability of the algorithm.

[0101] Furthermore, the optimal individual selection. Traditional single-objective whale optimization algorithms can select the optimal individual based on the value of the fitness function. However, a population with multiple objective functions often has multiple optimal individuals. What criteria should be used to select the leader whale group to update the position is one of the important issues in designing a multi-objective whale optimization algorithm.

[0102] Currently, most multi-objective optimization algorithms use the degree of crowding between individuals to rank the non-dominated solution sets. However, the calculation process is complex and not conducive to the algorithm's global search. Therefore, a method based on adaptive grid partitioning is used to select the optimal individual. Grid density can be used to describe the number of individuals contained in each region of the target space. The more individuals in a region, the greater the grid density; the fewer individuals, the smaller the grid density. To maintain the diversity of individuals in the population, individuals in regions with smaller grid density have a greater chance of being selected as the optimal individual. Taking objective function 1 as an example, the mathematical model for the grid information calculation method is shown in Equation (15).

[0103] Where: M represents the grid number of the individual; [·] represents the largest integer smaller than the number in the brackets; Grid is the number of grids to be divided; t is the number of iterations; y i (t) is the objective function value of individual i in the t-th generation; maxY(t) is the maximum value of the corresponding objective function among all individuals in the t-th generation population; minY(t) is the minimum value of the corresponding objective function among all individuals in the t-th generation population.

[0104] At this time, the grid number tuples of all individuals are obtained by formula (15), and the best individual is selected by the roulette method. The more individuals in the same number group, the greater the density of the grid, and the smaller the probability of selecting an individual in this grid as the best individual. When there are multiple individuals in the selected grid, one of them is randomly selected as the best individual.

[0105] The point is that swarm intelligence algorithms incorporate a Pareto external archive to store non-dominated solutions generated during the population iteration process. When the number of individuals in the external archive exceeds a pre-set threshold, a corresponding strategy is used to remove excess individuals to maintain the superiority of the individuals in the external archive. Following this principle, when the number of non-dominated solutions in the external archive exceeds a set threshold, the external archive can be processed based on the grid density of different individuals, removing individuals with high grid density from the archive until the required number of individuals in the external archive is met.

[0106] On the other hand, this embodiment also provides a multi-objective county distribution network optimization system, which includes:

[0107] The first division unit collects geographic information of the county area and performs preliminary division of the power supply grid.

[0108] The second division unit divides the power grid for the second time through the K-means clustering algorithm.

[0109] The third division unit selects the best planning strategy through the adaptive multi-objective whale optimization algorithm and divides the power supply grid for the third time.

[0110] If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program code.

[0111] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0112] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0113] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0114] Example 2

[0115] 3 , which is an embodiment of the present invention, provides a county-level distribution network grid planning method using resource digital fusion technology. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0116] The implementation steps of the multi-objective grid optimization algorithm are shown in Figure 3.

[0117] (1) Initialize the population fitness through the second grid division result.

[0118] (2) Update the Pareto profile and select the optimal individual.

[0119] (3) Update the positions of all whale individuals according to the formula.

[0120] (4) Calculate the fitness of the updated population.

[0121] (5) Store the non-dominated individuals in the updated population into the Pareto archive.

[0122] (6) If the number of individuals in the Pareto file exceeds the set maximum value, calculate the grid number of the individual, calculate the grid density, and delete the individuals with high grid density.

[0123] (7) Calculate the grid number and density of individuals in the Pareto archive.

[0124] (8) Use the roulette method to select the best individual.

[0125] (9) If the number of iterations reaches the set value, proceed to the next step. Otherwise, jump to step (3).

[0126] (10) Derive the Pareto optimal solution set as the third meshing result.

[0127] Through comparative testing of random scenes, the same scene was selected and divided using the present invention and the traditional method respectively. The test data is shown in Table 1.

[0128] As shown in Table 1, this invention improves the accuracy of power grid division, thereby ensuring efficient resource allocation and utilization. It also reduces unnecessary construction and maintenance costs, achieving higher economic benefits. The optimized grid design and resource allocation improve overall system reliability and reduce the frequency of failures. It supports rapid adjustment and expansion, enabling the distribution network to more flexibly respond to future changes and demands. Optimized grid planning reduces energy consumption and carbon emissions during construction and operation, contributing to environmental protection and sustainable development.

[0129] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A county-level distribution network grid planning method based on resource digital fusion technology is characterized by: include: Collect geographic information of county areas and conduct preliminary division of power supply grids; The power grid is divided for the second time using the K-means clustering algorithm; The adaptive multi-objective whale optimization algorithm is used to select the best planning strategy and perform the third division of the power grid.

2. The multi-objective county distribution network optimization method according to claim 1, characterized in that: The preliminary division includes the first division of the power supply grid on the power grid digital platform according to the geographical conditions of mountains, waters, transportation, and different functional areas within the county.

3. The multi-objective county distribution network optimization method according to claim 1, characterized in that: The K-means clustering algorithm includes determining the number of clusters k in the county area based on the preliminary grid division results, and using the geometric center of the grid as the center point of each cluster; Determine the classification conditions, and use the power supply distance between the substation and the load point and the distributed resources as the classification conditions; Based on the measurement of power supply range, each node is compared with the nearest cluster center and assigned to the cluster with the closest distance; Recalculate the data points in each cluster, that is, the mean of the node objects in each cluster, and calculate the clustering results of all clusters at the same time; If the result value changes, repeat the steps of determining the classification conditions, otherwise the algorithm ends and outputs the second mesh division result.

4. The multi-objective county distribution network optimization method according to claim 3, characterized in that: The third division includes establishing a grid-based multi-objective coordinated planning model for the county distribution network, and achieving the third division by solving the multi-objective optimization model; The county-level distribution network grid-based multi-objective coordinated planning model includes the following: total grid planning investment cost: Among them, C eco represents the total investment cost of grid planning; ε=k1+k2, where k1 and k2 are the operation and maintenance cost coefficient and the depreciation investment recovery coefficient respectively; C ij,m represents the total cost of the m-th inter-station power supply grid trunk line, which is the comprehensive cost of the entire length of all trunk lines; C ih,n represents the comprehensive cost of the trunk line of the self-loop power supply unit in the nth non-inter-station power supply grid; C fs,n represents the comprehensive cost of the trunk line of the radial power supply unit in the nth non-inter-station power supply grid; represents the annual cost of power loss on the main line of the m-th inter-station power supply grid; represents the annual cost of power outage loss on the trunk line of the nth non-inter-station power supply grid; C in represents the investment and construction cost of distributed resources; C op represents the operation and maintenance cost of distributed resources; C pro represents the distributed resource benefit; C l represents the investment in digital communication between grids; Assume that the interconnection line is built along the power line, and each node has a digital interconnection substation, and the interconnection substation is not far from the substation; set the digital interconnection investment to: Where N represents the number of grids, d m represents the branch line length of grid m, p l Indicates the unit price of data transmission line, p c Indicates the unit price of the grid control unit; The power outage loss describes the overall reliability index of the distribution network and is expressed as: Among them, C los It is the reliability index of the distribution network, indicating the total power outage loss of the entire network; P lo,m,j represents the power outage probability of the jth node in the grid m; N los represents the number of power outages in the distribution network during the planning period; t represents the duration of a single power outage; N represents the number of load nodes in each grid; S m,j is the power supply status of load node j in grid m, 0 means power failure, 1 means normal; P m,j (t) is the actual power consumed by load node j in grid m at time t; P m,j (t)=P den,m,j (t)-P dis,m,j (t) Among them, P den,m,j (t) represents the power demand of load node j in grid m at time t; P dis,m,j (t) represents the power generated by the distributed resources of load node j in grid m at time t; Set the inter-grid power interaction objective function: Among them, P ij represents the active power flow of branch i between grids, Q ij represents the reactive power flow of inter-grid branch i, k represents the number of inter-grid branches, P p 、P q represents the power interaction coefficient.

5. The multi-objective county distribution network optimization method according to claim 4, characterized in that: The multi-objective optimization model solution includes: the mathematical model of the multi-objective optimization problem is expressed as: min F(x)=(C eco ,C los ,E m,n ) Introducing the Pareto dominance relationship, for m objective functions f i (x), i=1,2,...,m Given any two decision variables x1 and x2, if they satisfy the following equation, then the solution x1 is said to dominate x2; The whale's hunting strategy is a process of continuously approaching the optimal solution to the optimization problem, where each whale can be regarded as a solution to the optimization problem, and the prey can be regarded as the optimal solution to be found. The optimization problem is solved by imitating the whale's hunting process. Surrounding the prey: In the optimization problem, the individual with the best fitness function is regarded as the optimal individual in the population. The process of other individuals in the population changing their positions towards the optimal individual is expressed by the mathematical model as follows: Where t represents the current number of iterations, Indicates the current position of the whale, Indicates the currently obtained optimal whale position, Indicates the whale's position after the updated position. represents the distance the whale moves, A and C are the correlation coefficients; A=2aR1-a,C=2R2,a=2(t max -t) / t max Among them, R1 and R2 represent random floating point numbers on [0, 1], t max Indicates the maximum number of iterations set, and a indicates that the convergence factor will gradually decrease to 0 as the number of iterations increases; Bubble net hunting involves spiral swimming: whales will swim in spirals towards their prey, which can be expressed mathematically as follows: in, represents the distance between the whale and the optimal individual; b represents a constant used to define the spiral shape, usually 1; l is a random floating point number on (-1, 1).

6. The multi-objective county distribution network optimization method according to claim 5, characterized in that: The hunting strategy also includes introducing an action probability P to determine whether the whale chooses to approach the prey by shrinking and encircling or swimming in a spiral. When the generated random number is less than P, the whale will shrink and encircle, and when it is not less than P, it will spiral. The mathematical model of whale bubble net hunting is: Where p represents a random floating point number in the interval [0, 1]; P represents the action probability constant, which is usually set to 0.5; Search hunting: Whales search for prey globally, and its mathematical model is expressed as: in, represents the position of a randomly selected whale in the population; When A≥1, the whale randomly selects individuals in the population to update their positions, enhancing the global search capability of the algorithm.

7. The multi-objective county distribution network optimization method according to claim 6, characterized in that: The mathematical model for grid information calculation is expressed as: Where M represents the grid number of the individual, [·] represents the largest integer smaller than the number in the brackets, Grid represents the number of grids to be divided, t represents the number of iterations, and y i (t) represents the objective function value of individual i in the t-th generation, max Y(t) represents the maximum value of the corresponding objective function among all individuals in the t-th generation population, and min Y(t) represents the minimum value of the corresponding objective function among all individuals in the t-th generation population; Get the grid number tuples of all individuals, use the roulette wheel method to select the best individual, and randomly select one of the individuals as the best individual when the selected grid has multiple individuals.

8. A multi-objective county distribution network optimization system using the method according to any one of claims 1 to 7, characterized in that: The first division unit collects geographic information of the county area and performs preliminary division of the power supply grid; The second division unit divides the power grid for the second time through the K-means clustering algorithm; The third division unit selects the best planning strategy through the adaptive multi-objective whale optimization algorithm and divides the power supply grid for the third time.

9. A computer device comprising: memory and processor; The memory stores a computer program, characterized in that when the processor executes the computer program, the steps of the multi-objective county distribution network grid optimization method are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the multi-objective county distribution network grid optimization method are implemented.

Images