Method for realizing power distribution network reconstruction by using flower pollination algorithm optimized by chaos golden search

Abstract

The invention relates to a method for realizing power distribution network reconstruction by using a flower pollination algorithm optimized by chaos golden search, which belongs to the field of power distribution network optimization operation and comprises the following steps of: establishing an objective function and determining system power loss; reading data of a test system, including data of a branch circuit and a bus, and calculating power loss; taking minimization of power loss as a main target, and calculating optimal configuration based on an improved flower pollination algorithm; checking the satisfaction degree of the constraints of the power system, the power system must meet certain constraint conditions, if the constraints conflict, repeating the process, otherwise, storing the optimal configuration result, and stopping the process; the method has the advantages of being short in convergence time, high in convergence speed, capable of guaranteeing convergence and easy to implement, and can meet the requirements of multivariable and non-convex constraint large-scale optimal reconstruction problems.

Description

Technical Field

[0001] The invention belongs to the field of optimized operation of power distribution networks, and in particular relates to a method for realizing power distribution network reconstruction by using a flower pollination algorithm optimized by chaotic golden search. Background Art

[0002] The distribution network is the last link of the power grid in distributing electric energy. It is directly facing the user load. It involves the distribution of electric energy and the power supply on the user side. Therefore, it is an important part for users to evaluate the performance of the entire power system and the quality of user experience. At present, power loss is one of the most important issues in the distribution network. However, in the distribution stage, since low-voltage and high-current account for more than 13% of the total power generation, the control of power loss will be more important. In order to solve this problem, many solutions have been proposed, such as distributed flexible AC transmission equipment installed in the distribution network, distributed generation units or capacitors, and distribution network reconstruction. Among them, distribution network reconstruction is considered to be the most economical solution, and in terms of sustainability, distribution network reconstruction is also an effective measure to reduce power loss. It is a process of changing the network topology by changing the closing and closing states of the tie lines, segmenting the system, and making the new configuration more suitable and effective, under the premise of satisfying the radial network configuration, power balance, bus voltage and branch current limit and other equations and inequality constraints, so that the new configuration is more suitable and more effective. However, it is a major challenge to obtain minimized power loss and maintain voltage distribution through the reconfiguration of the distribution network. Since there are many possible exchange combinations for the distribution network, each of which is modeled based on a binary variable, a complex combinatorial optimization problem is proposed. In addition, the power flow equations are non-convex and have nonlinear constraints. Therefore, the entire optimization problem is a non-convex combinatorial problem. Establishing the optimal distribution network reconstruction requires the application of global optimization methods to deal with large-scale problems with non-convexity in the constraints. One of the best strategies to solve non-convex problems is to use meta-heuristic optimization algorithms, which are otherwise challenging to use other techniques. These algorithms are usually easy to apply, do not require objective function gradients, and often find an adequate solution. However, the performance and convergence speed of these algorithms vary depending on the size of the problem and the constraints. Therefore, finding the best algorithm for the optimal reconstruction has received considerable attention in power system research. To address this problem, a variety of meta-heuristic optimization algorithms have been used, including genetic algorithms, particle swarm optimization algorithms, ant colony optimization algorithms, butterfly optimization algorithms, gravitational search algorithms, cuckoo search algorithms, artificial bee colony algorithms, teaching-based optimization algorithms, flower pollination algorithms, etc. At present, these algorithms are less efficient. In addition to their relatively long simulation time, they are also not guaranteed to locate the global optimal solution. Therefore, they may get stuck in local optima. This shortcoming becomes increasingly apparent when applied to large-scale problems with many variables and non-convex constraints.

[0003] Recent research has taken a step forward and focused on integrating optimization algorithms to build new hybrid algorithms that can benefit from the specific strengths of individual algorithms and mitigate their inherent weaknesses. These integrated algorithms often show better performance than individual algorithms or improved meta-heuristic optimization algorithms when solving the optimal distribution network reconstruction problem. They usually require fewer iterations to converge and can find better solutions in lower simulation time. However, there is still a research gap on how to solve large-scale optimal reconstruction problems with so many continuous and integer variables in a reasonable time. Summary of the invention

[0004] The technical problem to be solved by the present invention is to propose a distribution network reconstruction optimization method based on an improved flower pollination algorithm, which involves the integration of a flower pollination algorithm (FPA) and a chaotic golden search algorithm (CGSA), and has the advantages of short convergence time, fast convergence speed, guaranteed convergence, and easy implementation, and can meet the needs of large-scale optimal reconstruction problems with multi-variables and non-convex constraints.

[0005] The technical solution of the present invention comprises the following steps:

[0006] Step 1, establishment of objective function and determination of system power loss;

[0007] Step 2, reading the data of the test system, including the data of the branch lines and the bus, and calculating the power loss;

[0008] Step 3, with minimizing power loss as the main goal, the optimal configuration is calculated based on the improved flower pollination algorithm;

[0009] Step 4: Check the satisfaction degree of power system constraints. The power system must meet certain constraints. If there is a conflict in the constraints, this process will be repeated. Otherwise, the optimal configuration result will be saved and the process will stop.

[0010] Furthermore, step 1 is specifically as follows:

[0011] 1.1, with the goal of reducing the power loss of the system, the objective function is:

[0012] OBJ F =MIN(P Loss )

[0013] Where P Loss is the power loss, OBJ F is the objective function;

[0014] 1.2, determine the power loss in the circuit by forward scanning and backward scanning, the mathematical expression of power loss is:

[0015]

[0016] Where N BR is the total number of distribution network branches, is the voltage amplitude of busbar I, is the active load of busbar I, is the reactive load of busbar I, R I is the resistance of branch I.

[0017] Furthermore, step 2 is specifically as follows:

[0018] For a decision vector with a set of switch quantities, distributed generation and capacitor reactive power, the branches corresponding to the switch groups are cut out, and the equivalent reactive power of the capacitors is subtracted from the reactive power demand of these buses to calculate the equivalent reactive demand of the bus containing the capacitors; the equivalent active demand of each bus is calculated by subtracting the active power of the dispatchable and renewable DGs connected to the bus from its active demand. Then, all nodes of the system should satisfy the following set of nonlinear equations:

[0019]

[0020] Where P gi and Q gi are the actual active power and reactive power of the ith bus, P li and Q li are respectively the active and reactive demands of the ith bus; B ij and Q ij Respectively represent the real and imaginary parts of the i-th and j-th elements in the node admittance matrix of the system; δ i and δ j Represent the voltage angles of the i-th and j-th buses respectively;

[0021] Considering the effects of reconstruction, distributed power sources and capacitors, the Newton-Raphson algorithm is used to solve the power flow equation, calculate the bus voltage, and then calculate the power loss using the following formula:

[0022]

[0023] Where l represents the number of branches, i and j represent the busbars connected to the branch, and R l and X l Respectively represent the resistance and reactance of the lth branch, N l Represents the total number of branches in the system, V i and V j Represents the voltage phasor of the connected bus.

[0024] Furthermore, step 3 is specifically as follows:

[0025] Apply the chaotic golden search algorithm to local pollination, use the chaotic sequence and the golden ratio to enhance the local pollination process, and optimize the local search of the flower pollination algorithm using a new scaling factor to obtain the best performance. The main goal of the chaotic golden search algorithm is to find the possible minimum value within the interval by introducing a chaotic sequence. The introduction of the chaotic sequence can increase the diversity and global search ability of the algorithm;

[0026] The chaotic golden ratio is calculated by the following formula:

[0027]

[0028] The generation of the chaotic sequence can be achieved through the Logistic map. The formula for generating the chaotic sequence using the Logistic map is:

[0029] x n+1 = rx n (1 - x n )

[0030] where r is the control parameter, x n is the nth value in the sequence, and the initial value x0 can be randomly selected. The main goal of the chaotic golden search algorithm is to find the possible minimum F(X) within the interval [l, u], X ∈ R; the two boundary points X1, X2 ∈ [l, u] are calculated by the following equations:

[0031] X1 = l + (1 - C)(l - u)

[0032] X2 = l + C(l - u)

[0033] where X1 and X2 are the boundaries of the unimodal optimization curve, the upper and lower limits of the interval points are represented by l and u, and C is the golden ratio; the objective function is evaluated at these two points as the functions f(X1) and f(X2) respectively. If f(X1) < f(X2), the optimal solution is related to [l, x1], otherwise, the optimal solution is related to [x2, u]. Repeat this process continuously and introduce a chaotic sequence to adjust the search interval until the termination condition is met.

[0034] Furthermore, the improved flower pollination algorithm is specifically as follows:

[0035] Step 3.1, Initialize the algorithm parameters, including the initial population size of pollen, the conversion probability, the initial value of the chaotic sequence, and the maximum number of iterations of the algorithm;

[0036] Step 3.2, Generate the initial population, generate the initial population through the initial population generation rule and screen the feasible pollen;

[0037] Step 3.3, Calculate the fitness values of each feasible pollen, and find the optimal solution g * ;

[0038] Step 3.4, enter the main loop, and generate new pollen through global pollination or local pollination according to the conversion probability;

[0039] Step 3.5, determine whether the individual needs to be updated and calculate the fitness value of the newly generated pollen;

[0040] Step 3.6, update the optimal pollen;

[0041] Step 3.7, determine whether the termination condition is met, if so, the algorithm terminates; otherwise, return to step 3.2; repeat steps 3.2-3.6.

[0042] The step 3.4, wherein

[0043] Global pollination: Insects fly with different strides and move long distances. Levy flight is a mobile scenario that can better simulate motion characteristics. Therefore, L>0 is considered to be a Levy distribution, which is expressed by the following formula:

[0044]

[0045] Where Γ(λ) is the standard gamma function, and this distribution is valid for larger step sizes M>0; the new population is generated by the following formula:

[0046]

[0047] Where L(λ) is the flight step length related to pollination intensity, γ is the proportional factor for adjusting the step length, and g * is the current optimal solution, is the pollen at iteration T, It is the pollen at iteration T+1;

[0048] Local pollination: Local pollination only occurs in a small part of the current population, so the uniform distribution determines the step size, and the new population is generated by the following formula:

[0049]

[0050] in, and Pollen from different flowers of the same plant, ε is a proportional coefficient that follows a uniform distribution; this method optimizes the local search by using a new proportional coefficient to obtain the best performance; and ultimately generates two intermediate points, as shown in the following formula:

[0051]

[0052] Furthermore, step 4 is specifically as follows:

[0053] 4.1, Feeder capacity limitation,

[0054] In the power system, the branch current flow should be within the maximum limit, as shown in the following formula:

[0055] 0≤I I ≤I MAX ; I=1,...,N BR

[0056] In the formula, I MAX is the maximum current in the branch of the system, I I is the current flowing in branch I;

[0057] 4.2, bus voltage limit,

[0058] In the power system, the bus voltage must follow the maximum and minimum values, as shown in the following formula:

[0059] V MIN,I ≤V I ≤V MAX，I ; I=1,...,N B

[0060] Where V MAX,I is the maximum voltage of busbar I, V MIN,I is the minimum voltage of busbar I;

[0061] 4.3, real power and reactive power balance, in the power system, the active power and reactive power should be within their respective limits, as shown in the following formula:

[0062]

[0063] Where P L,K is the active power loss of the Kth branch, Q L,K is the reactive power loss of the Kth branch, P D,J is the active load demand of busbar J, Q D,J is the reactive load demand of busbar J, P DG,I is the actual output power of the Ith DG unit, Q DG,I is the reactive power output of the first distributed generation, P SLACK is the active power provided by the standby bus, Q SLACK is the reactive power provided by the standby bus, N B is the total number of buses in the distribution network, N DG is the total number of distributed generation in the system;

[0064] 4.4: Radial configuration constraints,

[0065] The number of branch and tie switches is determined by the following formula:

[0066] N Br =(N B -1)

[0067]

[0068] Where N TS is the total number of tie lines in the distribution network, is the total number of lines in the ring network, is the total number of lines in the radial distribution system.

[0069] Beneficial effects of the present invention:

[0070] The present invention discloses a distribution network reconstruction optimization method using a flower pollination algorithm optimized by chaos golden search, and applies the chaos golden search algorithm to the local search of the flower pollination algorithm. The search interval is effectively narrowed by the chaos golden search algorithm, and then the flower pollination algorithm searches the updated interval to obtain the optimal solution, and can perform a deep search with a lower convergence time. When selecting the best feeder connection of the power system, the proposed method can quickly converge the system in a radial configuration system, calculate the best position of the feeder that reduces the power loss of the system, so as to minimize the power loss, and has significant advantages in reducing power loss. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 is a flow chart of the present invention;

[0072] Figure 2 IEEE 119 bus system diagram of the present invention;

[0073] Figure 3 A power loss comparison diagram of different algorithms in the IEEE 119 bus system of the present invention;

[0074] Figure 4 It is the normalized calculation time of the present invention in all scenarios in the IEEE 119-node system. DETAILED DESCRIPTION

[0075] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention are clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0076] Example

[0077] The present invention provides a distribution network reconstruction optimization method based on an improved flower pollination algorithm, comprising the following steps:

[0078] Step 1: Establish the objective function and determine the average power loss of the system.

[0079] 1.1, with the goal of reducing the power loss of the system, the objective function is:

[0080] OBJ F =MIN(P Loss )

[0081] Where P Loss is the power loss, OBJ F is the objective function.

[0082] 1.2, determine the power loss in the circuit by forward scanning and backward scanning. The mathematical expression of power loss is:

[0083]

[0084] Where N BR is the total number of distribution network branches, is the voltage amplitude of busbar I, is the active load of busbar I, is the reactive load of busbar I, R I is the resistance of branch I.

[0085] Step 2: Read the data of the test system, including the data of the branch lines and the bus, and calculate the power loss.

[0086] For a decision vector with a set of switch quantities, DG generation and capacitor reactive power, the branches corresponding to the switch groups are cut out, and the equivalent reactive power of the capacitors is subtracted from the reactive power demand of these buses to calculate the equivalent reactive power demand of the bus containing the capacitors. The equivalent active power demand of each bus is calculated by subtracting the active power of the dispatchable and renewable DGs connected to the bus from its active power demand. Then, all nodes of the system should satisfy the following set of nonlinear equations:

[0087]

[0088] Where P gi and Q gi are the actual active power and reactive power of the ith bus, P li and Q li are the active and reactive demands of the ith bus, respectively. ij and Q ij Respectively represent the real and imaginary parts of the i-th and j-th elements in the node admittance matrix of the system.i and δ j Represent the voltage angles of the i-th and j-th buses respectively.

[0089] Considering the effects of reconstruction, distributed power sources and capacitors, the Newton-Raphson algorithm is used to solve the power flow equation, calculate the bus voltage, and then calculate the power loss using the following formula.

[0090]

[0091] Where l represents the number of branches, i and j represent the busbars connected to the branch, and R l and X l Respectively represent the resistance and reactance of the lth branch, N l Indicates the total number of branches in the system. V i and V j Represents the voltage phasor of the connected bus.

[0092] Step 3: With minimizing power loss as the main goal, the optimal configuration is calculated based on the flower pollination algorithm optimized using chaotic golden search.

[0093] Improved flower pollination algorithm: The chaotic golden search algorithm is applied to local pollination, and the chaotic sequence and golden ratio are used to enhance the local pollination process. The local search of the flower pollination algorithm is optimized using a new proportional factor to obtain the best performance. The main goal of the chaotic golden search algorithm is to find the possible minimum value in the interval by introducing a chaotic sequence. The introduction of a chaotic sequence can increase the diversity of the algorithm and the global search capability.

[0094] The chaos golden ratio is calculated by the following formula:

[0095]

[0096] The generation of chaotic sequences can be achieved through Logistic mapping. The formula for generating chaotic sequences using Logistic mapping is:

[0097] x n+1 =rx n (1-x n )

[0098] Where r is the control parameter. n is the nth value in the sequence, and the initial value x0 can be randomly selected. The main goal of the chaotic golden search algorithm is to find the minimum possible F(X) in the interval [l,u], X∈R. The two boundary points X1, X2∈[l,u] can be calculated by the following formula:

[0099] X1=l+(1-C)(lu)

[0100] X2 = l + C(l - u)

[0101] Among them, X1 and X2 are the boundaries of the unimodal optimization curve. The upper and lower limits of the interval points are represented by l and u, and C is the golden ratio. The objective function is evaluated at these two points as functions f(X1) and f(X2) respectively. If f(X1) < f(X2), the optimal solution is related to [l, x1]; otherwise, the optimal solution is related to [x2, u]. This process is repeated continuously, and a chaotic sequence is introduced to adjust the search interval until the termination condition is met.

[0102] The improved flower pollination algorithm is specifically as follows:

[0103] Step 3.1: Initialize the algorithm parameters, including the initial population size of pollen, the conversion probability, the initial value of the chaotic sequence, and the maximum number of algorithm iterations;

[0104] Step 3.2: Generate the initial population, and generate the initial population and screen the feasible pollen through the initial population generation rule;

[0105] Step 3.3: Calculate the fitness values of each feasible pollen, and find the optimal solution g in the initial population * ;

[0106] Step 3.4: Enter the main loop, and generate new pollen through global pollination or local pollination according to the conversion probability.

[0107] Among them, for global pollination: Insects fly with different step sizes and move long distances. Levy flight is a movement scenario that can better simulate the movement characteristics. Therefore, L > 0 is considered to be the Levy distribution, which is expressed by the following formula:

[0108]

[0109] Among them, Γ(λ) is the standard gamma function, and this distribution is valid for larger step sizes M > 0.

[0110] The new population is generated by the following formula:

[0111]

[0112] In the formula, L(λ) is the flight step size related to the pollination intensity, γ is the proportionality factor for adjusting the step size, g * is the current optimal solution, is the pollen at iteration T, is the pollen at iteration T + 1.

[0113] For local pollination: Local pollination only occurs in a small part of the current population. Therefore, the uniform distribution determines the step size. The new population is generated by the following formula:

[0114]

[0115] in, and Pollen from different flowers of the same plant, ε is a proportional coefficient that follows a uniform distribution. This method uses a new proportional coefficient to optimize the local search to obtain the best performance. Finally, two intermediate points are generated, as shown in the following formula:

[0116]

[0117] Step 3.5: Determine whether the individual needs to be updated and calculate the fitness value of the newly generated pollen;

[0118] Step 3.6: Update the optimal pollen;

[0119] Step 3.7: Determine whether the termination condition is met. If so, the algorithm terminates; otherwise, return to step 3.2; repeat steps 3.2-3.6.

[0120] Step 4: Check the satisfaction of power system constraints. The power system constraints are related to the objective function of power loss. In order to maintain efficient power flow operation, the power system must meet certain constraints. If the constraints produce some conflicts, the process will be repeated. Otherwise, the optimal configuration result is saved and the process stops.

[0121] 4.1: Feeder capacity limitation

[0122] In the power system, the branch current flow should be within the maximum limit, as shown in the following formula:

[0123] 0≤I I ≤I MAX ; I=1,...,N BR

[0124] In the formula, I MAX is the maximum current in the branch of the system, I I is the current flowing in branch I.

[0125] 4.2: Bus voltage limit

[0126] In the power system, the bus voltage must follow the maximum and minimum values, as shown in the following formula:

[0127] V MIN,I ≤V I ≤V MAX，I ; I=1,...,N B

[0128] Where V MAX,I is the maximum voltage of busbar I, V MIN,Iis the minimum voltage of bus I.

[0129] 4.3: Real power and reactive power balance

[0130] In the power system, active power and reactive power should be within their respective limits, as shown in the following equation:

[0131]

[0132] Where P L,K is the active power loss of the Kth branch, Q L,K is the reactive power loss of the Kth branch, P D,J is the active load demand of busbar J, Q D,J is the reactive load demand of busbar J, P DG,I is the actual output power of the Ith DG unit, Q DG,I is the reactive power output of the first distributed generation unit, P SLACK is the active power provided by the standby bus, Q SLACK is the reactive power provided by the standby bus, N B is the total number of buses in the distribution network, N DG is the total number of distributed generation units in the system.

[0133] 4.4: Radial Configuration Constraints

[0134] In order to minimize the fault level while saving energy and protecting the equipment, the distribution network must be operated in radial networking. Therefore, it is necessary to determine the configuration of the radial network when dealing with the reconfiguration problem. The proposed method is feasible for radial configuration of the distribution network. The proposed method cannot be applied to meshed distribution systems because it will lead to negative short-circuit consequences. The number of branches and tie switches can be determined by the following formula:

[0135] N Br =(N B -1)

[0136]

[0137] Where N TS is the total number of tie lines in the distribution network, is the total number of lines in the ring network, is the total number of lines in the radial distribution system.

[0138] Experimental verification:

[0139] Relevant simulation and modeling work are carried out, and the method is compared and analyzed with existing FPA, CFPA, GSFA and other related algorithms. The PL of the system is reduced mainly by changing the connection mode of the tie switch. This experiment will test three main scenarios and verify the effectiveness of the proposed algorithm through application in the IEEE 119 bus system.

[0140] Scenario 1: Bus Fault (BF)

[0141] Scenario 2: Line Fault (LF)

[0142] Scenario 3: Bus and Line Failure (BLF)

[0143] Use as Figure 2 The IEEE119 bus system shown in the figure was tested for performance on three test cases. Three distributed generation units (DG) and one energy storage system (ESS) were placed on busbars 28, 78, 114 and 55 respectively. The dotted lines represent switches that are turned on or off for dynamic distribution of feed. The total reactive load and real power load of the system are 17.04 MVAr and 22.7097 M respectively. As shown in Table 1, the optimal tie line connection mode of the system is obtained. Figure 3 As shown, it is the percentage of power consumption saved in the bus system. In the case of line fault, the highest power consumption saving of the system is 15.89%. Similarly, in the case of bus fault (BF) and bus and line fault (BLF), the power loss is saved by 4.84% and 17.22%, respectively. The results show that the method proposed in the present invention has good performance and has great advantages in reducing power loss.

[0144] The above are only specific embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0145] Table 1 Optimal tie switch connection positions achieved by reconfiguration in IEEE119 busbar system

[0146]

Claims

1. A method for realizing distribution network reconstruction using a flower pollination algorithm optimized by chaotic golden search, characterized in that: It includes the following steps: Step 1, establishment of the objective function and determination of the system power loss; Step 2, read the data of the test system, including the data of branch lines and buses, and calculate the power loss; Step 3, with minimizing the power loss as the main goal, calculate the optimal configuration based on the improved flower pollination algorithm; Step 4, check the satisfaction degree of the power system constraints. The power system must meet certain constraint conditions. If conflicts occur in the constraints, this process will be repeated. Otherwise, save the optimal configuration result and the process stops.

2. According to claim 1, a method for realizing distribution network reconstruction using a flower pollination algorithm optimized by chaotic golden search is characterized in that: The specific content of the said Step 1 is as follows: 1.1, with the goal of reducing the power loss of the system, the objective function is: OBJ F =MIN(P Loss ) Where P Loss is the power loss, OBJ F is the objective function; 1.2, determine the power loss situation in the circuit through forward scanning and backward scanning. The mathematical expression of the power loss is: Where N BR is the total number of distribution network branches, is the voltage amplitude of busbar I, is the active load of busbar I, is the reactive load of busbar I, R I is the resistance of branch I.

3. According to claim 1, a method for realizing distribution network reconstruction using a flower pollination algorithm optimized by chaotic golden search is characterized in that: The specific content of the said Step 2 is as follows: For a decision vector with a certain set of switch quantities, distributed power generation and capacitor reactive power, cut out the branches corresponding to this set of switch groups, subtract the equivalent reactive power of the capacitor from the reactive power demands of these buses, and then the equivalent reactive power demand of the buses containing capacitors can be calculated; the equivalent active power demand of each bus is calculated by subtracting the active power of the dispatchable and renewable DGs connected to this bus from its active power demand. Then, all nodes of the system should satisfy the following set of nonlinear equations: Where P gi and Q gi are the actual active power and reactive power of the ith bus, P li and Q li are respectively the active and reactive demands of the ith bus; B ij and Q ij Respectively represent the real and imaginary parts of the i-th and j-th elements in the node admittance matrix of the system; δ i and δ j Represent the voltage angles of the i-th and j-th buses respectively; Considering the influence of reconstruction, distributed power sources and capacitors, use the Newton-Raphson algorithm to solve the power flow equation, calculate the bus voltage, and then calculate the power loss through the following formula Where l represents the number of branches, i and j represent the busbars connected to the branch, and R l and X l Respectively represent the resistance and reactance of the lth branch, N l Represents the total number of branches in the system, V i and V j Represents the voltage phasor of the connected bus.

4. According to claim 1, a method for realizing distribution network reconstruction using a flower pollination algorithm optimized by chaotic golden search is characterized in that: The specific content of the said Step 3 is as follows: Apply the chaotic golden search algorithm to local pollination, use the chaotic sequence and the golden ratio to enhance the local pollination process, and optimize the local search of the flower pollination algorithm with a new scaling factor to obtain the best performance. The main goal of the chaotic golden search algorithm is to find the possible minimum value within the interval by introducing the chaotic sequence. The introduction of the chaotic sequence can increase the diversity and global search ability of the algorithm; The chaotic golden ratio is calculated by the following formula: The generation of the chaotic sequence can be achieved through the Logistic map. The formula for generating the chaotic sequence using the Logistic map is: x n+1 =rx n (1-x n ) Among them, r is the control parameter, x n is the nth value in the sequence, and the initial value x0 can be randomly selected. The main goal of the chaotic golden search algorithm is to find the minimum possible F(X) in the interval [l,u], X∈R; the two boundary points X1, X2∈[l,u] are calculated by the following formula: X1 = l + (1 - C)(l - u) X2 = l + C(l - u) Where, X1 and X2 are the boundaries of the unimodal optimization curve. The upper and lower limits of the interval points are represented by l and u, and C is the golden ratio; the objective function is evaluated as the function f(X1) and f(X2) at these two points respectively. If f(X1) < f(X2), the optimal solution is related to [l, x1], otherwise, the optimal solution is related to [x2, u]. Repeat this process continuously and introduce the chaotic sequence to adjust the search interval until the termination condition is met.

5. A method for realizing distribution network reconstruction using a flower pollination algorithm optimized by chaotic golden search according to claim 1 or 4, characterized in that: In the said Step 3, the improved flower pollination algorithm is specifically as follows: Step 3.1, initialize the algorithm parameters, the initial population size of pollen, the conversion probability, the initial value of the chaotic sequence, and the maximum number of algorithm iterations; Step 3.2, generate the initial population, generate the initial population through the initial population generation rule and screen the feasible pollen; Step 3.3, calculate the fitness value of each feasible pollen and find the optimal solution g in the initial population * ; Step 3.4, enter the main loop, generate new pollen through global pollination or local pollination according to the conversion probability; Step 3.5, determine whether the individual needs to be updated and calculate the fitness value of the newly generated pollen; Step 3.6, update the optimal pollen; Step 3.7, determine whether the termination condition is met, if so, the algorithm terminates; otherwise, return to step 3.2; repeat steps 3.2-3.

6.

6. According to claim 5, a method for realizing distribution network reconstruction using a flower pollination algorithm optimized by chaotic golden search is characterized in that: The step 3.4, wherein Global pollination: Insects fly with different strides and move long distances. Levy flight is a mobile scenario that can better simulate motion characteristics. Therefore, L>0 is considered to be a Levy distribution, which is expressed by the following formula: Where Γ(λ) is the standard gamma function, and this distribution is valid for larger step sizes M>0; the new population is generated by the following formula: Where L(λ) is the flight step length related to pollination intensity, γ is the proportional factor for adjusting the step length, and g * is the current optimal solution, is the pollen at iteration T, It is the pollen at iteration T+1; Local pollination: Local pollination only occurs in a small part of the current population, so the uniform distribution determines the step size, and the new population is generated by the following formula: in, and Pollen from different flowers of the same plant, ε is a proportional coefficient that follows a uniform distribution; this method optimizes the local search by using a new proportional coefficient to obtain the best performance; and ultimately generates two intermediate points, as shown in the following formula: 。 7. According to claim 1, a method for realizing distribution network reconstruction using a flower pollination algorithm optimized by chaotic golden search, characterized in that: The step 4 is specifically as follows: 4.1, Feeder capacity limitation, In the power system, the branch current flow should be within the maximum limit, as shown in the following formula: 0≤I I ≤I MAX ;I=1,...,N BR In the formula, I MAX is the maximum current in the branch of the system, I I is the current flowing in branch I; 4.2, bus voltage limit, In the power system, the bus voltage must follow the maximum and minimum values, as shown in the following formula: V MIN,I ≤V I ≤V MAX，I ;I=1,...,N B Where V MAX,I is the maximum voltage of busbar I, V MIN,I is the minimum voltage of busbar I; 4.3, real power and reactive power balance, In the power system, active power and reactive power should be within their respective limits, as shown in the following equation: Where P L,K is the active power loss of the Kth branch, Q L,K is the reactive power loss of the Kth branch, P D,J is the active load demand of busbar J, Q D,J is the reactive load demand of busbar J, P DG,I is the actual output power of the Ith DG unit, Q DG,I is the reactive power output of the first distributed generation, P SLACK is the active power provided by the standby bus, Q SLACK is the reactive power provided by the standby bus, N B is the total number of buses in the distribution network, N DG is the total number of distributed generation in the system; 4.4: Radial configuration constraints, The number of branch and tie switches is determined by the following formula: N Br =(N B -1) Where N TS is the total number of tie lines in the distribution network, is the total number of lines in the ring network, is the total number of lines in the radial distribution system.

Images