Distributed photovoltaic optimal configuration method based on improved multi-target circulation system algorithm

Abstract

The invention discloses a distributed photovoltaic optimal configuration method based on an improved multi-target circulation system algorithm. The distributed photovoltaic optimal configuration method is carried out through six steps. The distributed photovoltaic optimal configuration method has the following beneficial effects: firstly, through an established transformer area operation model and a photovoltaic system model, voltage fluctuation, a line loss rate and various operation constraint conditions are considered, and an improved multi-target CSBO algorithm is used for solving, so that a reasonable access position and rated capacity of the DPV can be determined, the line loss rate of the system is effectively reduced, and the system reliability is improved; and voltage distribution is improved. Secondly, the distributed photovoltaic access position and the rated capacity are obtained, and the maximum photovoltaic permeability which can be accessed in the transformer area can be obtained according to the optimal solution; and thirdly, the CSBO algorithm is improved by adopting a rapid non-dominated sorting method based on elite strategy selection, so that when the multi-objective optimization problem is processed, the population can be screened more reasonably, the DPV optimization configuration problem of the low-voltage transformer area can be solved more effectively, and the solving precision and efficiency of the algorithm are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of distributed photovoltaic technology, and particularly relates to a distributed photovoltaic optimization configuration method based on an improved multi-objective circulatory system algorithm. Background Art

[0002] Regarding the connection point and capacity selection problems of the distributed photovoltaic (DPV) power station in the low-voltage distribution network around the substation area, an optimization configuration model is established, and a method for simulating and evaluating the photovoltaic accommodation capacity of the distribution network and selecting the optimal accommodation scheme is proposed.

[0003] Existing research mostly focuses on the configuration capacity of DPV in medium-voltage and high-voltage distribution networks, and pays insufficient attention to the configuration of DPV in low-voltage substation areas. The low-voltage substation area directly faces users, and its power consumption stability is crucial. Moreover, there is little research on the maximum penetration rate of configurable DPV in low-voltage substation areas, which cannot provide sufficient technical support for actual projects.

[0004] At present, the circulatory system based optimization (CSBO) algorithm is an intelligent optimization algorithm proposed based on the functional model of the human blood circulatory system. As Figure 1 shown, it has the advantage of obtaining optimal and reasonable results with lower control parameters and performs well in avoiding falling into local optima. In the circulatory system algorithm, pulmonary circulation (weak individual optimization) and systemic circulation (strong individual optimization) are modeled as two independent groups with two different optimization periods. At each iteration, the population needs to be sorted, and the weaker population enters the pulmonary circulation, while the stronger population enters the systemic circulation. In the circulatory system algorithm, the population sorting is performed by descending order of the single objective function value, and the adaptability is poor.

[0005] Therefore, in view of the deficiencies of the existing technology, it is very necessary to provide a distributed photovoltaic optimization configuration method based on an improved multi-objective circulatory system algorithm to solve the deficiencies of the existing technology. Summary of the Invention

[0006] The purpose of the present invention is to provide a distributed photovoltaic optimization configuration method based on an improved multi-objective circulatory system algorithm to avoid the deficiencies of the existing technology. The distributed photovoltaic optimization configuration method based on the improved multi-objective circulatory system algorithm can improve the power supply stability of the low-voltage substation area, determine the maximum photovoltaic penetration rate, and has strong adaptability of the optimization algorithm.

[0007] The above object of the present invention is achieved by the following technical measures:

[0008] Provide a distributed photovoltaic optimization configuration method based on an improved multi-objective circulatory system algorithm, which is carried out by the following steps:

[0009] S1. Input the input conditions into the CSBO algorithm. The input conditions are low-voltage distribution network parameters, distributed photovoltaic characteristic parameters, and load data.

[0010] S2. Generate an initial population. The population consists of multiple individuals, each individual is a Pareto solution, a Pareto solution is a decision variable, and each decision variable includes the distributed photovoltaic access location and rated capacity.

[0011] S3. Simultaneously use fast non-dominated sorting and elitist strategy selection to screen the population and obtain the screened population.

[0012] S4. According to the objective value constraints, perform population evolution on the screened population in the CSBO algorithm to obtain a new evolved population, and determine whether the iteration termination condition is reached. When yes, define the previous population as the final population and enter S5; when no, define the new population as the population and return to S3.

[0013] S5. Select a Pareto solution from the corresponding Pareto solution set in the final population, and define the selected Pareto solution as the optimal solution.

[0014] S6. Perform cyclic superposition processing on the optimal solution in S5 according to the operation constraint conditions to obtain the maximum DPV penetration rate.

[0015] Preferably, the above S3 is specifically as follows:

[0016] S3.1. Classify the population into different non-dominated fronts through the fast non-dominated sorting method to obtain non-dominated sets at different levels, and enter S3.2.

[0017] S3.2. Screen out the optimal population from the non-dominated sets at different levels through the elitist strategy selection to obtain the screened population.

[0018] Preferably, the above S3.1 is specifically as follows: Define the number of individuals in the previous population as N, then the individuals in the previous population are respectively updated through the blood flow model in the CSBO algorithm, through the systemic circulation in the CSBO algorithm, and through the pulmonary circulation in the CSBO algorithm, and all are merged to obtain a population Rt with the number of individuals being 2N. Then, perform fast non-dominated sorting on the population Rt to obtain multiple non-dominated sets, and sort all non-dominated sets from the best to the worst in sequence, which are P1, P2,..., Pi.

[0019] Preferably, the above S3.2 is carried out by the following steps:

[0020] S3.2.1. Directly add the non-dominated set P1 to the new population Qt. When the new population Qt is less than N, proceed to S3.2.2; when the new population Qt is greater than N, select and add according to the crowding degree comparison method to make the new population Qt equal to N and then proceed to S3.2.4; when the new population Qt is equal to N, proceed to S3.2.4;

[0021] S3.2.2. Sequentially add the subsequent non-dominated sets to the new population Qt from the front to the back. After each addition, proceed to S3.2.3;

[0022] S3.2.3. When the new population Qt is equal to N, stop adding and proceed to S3.2.4; when the new population Qt is greater than N, select and add according to the crowding degree comparison method to make the new population Qt equal to N and then proceed to S3.2.4; when the new population Qt is less than N, return to S3.2.2;

[0023] S3.2.4. Define the current population Qt as the screened population.

[0024] Preferably, the above target values are the voltage fluctuations of the distribution network substation area obtained through power flow calculation in power system analysis and the line loss rate of the distribution network substation area obtained through power flow calculation in power system analysis.

[0025] Preferably, the above S6 is specifically as follows:

[0026] S6.1. Set the added value as A, and A > 0 kw;

[0027] S6.2. Conduct power flow calculation on the superposition of the added value in the optimal solution to obtain the superposition value, and then proceed to S6.3;

[0028] S6.3. Determine whether the superposition value exceeds the operating constraint conditions. If not, proceed to S6.4; if so, proceed to S6.5;

[0029] S6.4. Let the superposition value be the optimal solution and return to S6.2;

[0030] S6.5. Stop superposition and calculate the maximum DPV penetration rate based on the current superposition value.

[0031] Preferably, the above low-voltage substation area grid parameters are at least one of line resistance, reactance, node voltage level, or transformer parameters.

[0032] Preferably, the above distributed photovoltaic characteristic parameters are at least one of the power characteristics of photovoltaic cells, conversion efficiency, or the relationship between light intensity and power output.

[0033] Preferably, the above load data is at least one of the load magnitudes or load curves of each node in the low-voltage substation area.

[0034] Preferably, the above operating constraint conditions are at least one of upper and lower voltage constraints, power balance constraints, or line capacity constraints.

[0035] Preferably, the above iteration termination condition is the maximum number of iterations.

[0036] A distributed photovoltaic optimization configuration method based on an improved multi-objective cyclic system algorithm of the present invention is carried out by the following steps: S1. Input the input conditions into the CSBO algorithm, where the input conditions are low-voltage substation grid parameters, distributed photovoltaic characteristic parameters, and load data; S2. Generate an initial population, the population consists of multiple individuals, each individual is a Pareto solution, a Pareto solution is a decision variable, and each decision variable includes the distributed photovoltaic access location and rated capacity; S3. Simultaneously use fast non-dominated sorting and elitist strategy selection to screen the population to obtain the screened population; S4. According to the objective value constraint and evolve the screened population in the CSBO algorithm to obtain a new evolved population, and determine whether the iteration termination condition is reached. When it is, define the previous population as the final population and enter S5; when it is not, define the new population as the population and return to S3; S5. Select a Pareto solution from the corresponding Pareto solution set in the newly evolved population, and define the selected Pareto solution as the optimal solution; S6. Perform cyclic superposition processing on the optimal solution in S5 according to the operating constraint conditions to obtain the maximum DPV penetration rate. The beneficial effects of this distributed photovoltaic optimization configuration method are as follows: First, through the established substation operation model and photovoltaic system model, considering voltage fluctuations, line loss rate, and various operating constraint conditions, and using the improved multi-objective CSBO algorithm to solve, it can determine the reasonable access location and rated capacity of DPV, effectively reduce the system line loss rate, improve the voltage distribution, and enhance the power supply stability of the low-voltage substation area. Second, this distributed photovoltaic optimization configuration method obtains the distributed photovoltaic access location and rated capacity corresponding to the optimal solution, and can also obtain the maximum photovoltaic penetration rate that can be accessed in the substation area according to the optimal solution; the present invention provides a key reference for the subsequent investment and construction of photovoltaic power stations, helps to reasonably plan the scale of photovoltaic power stations, and improves the utilization efficiency of new energy. The prior art lacks research on the maximum penetration rate of DPV that can be configured in the low-voltage substation area, and the present invention makes up for this deficiency. Third, the present invention improves the CSBO algorithm by using the fast non-dominated sorting method based on elitist strategy selection. When dealing with multi-objective optimization problems, compared with the single-objective sorting method of the basic CSBO algorithm, it can more reasonably screen the population, more effectively solve the DPV optimization configuration problem in the low-voltage substation area, improve the solution accuracy and efficiency of the algorithm, and has the advantage of strong adaptability of the optimization algorithm. Description of the Drawings

[0037] The present invention will be further described with reference to the accompanying drawings, but the content in the drawings does not constitute any limitation to the present invention.

[0038] Figure 1 It is a flowchart of the existing CSBO algorithm at present.

[0039] Figure 2 In the present invention, it is a flowchart of the distributed photovoltaic optimization configuration method S1 - S4 for improving the multi - objective cyclic system algorithm.

[0040] Figure 3 It is a schematic diagram of elite strategy selection. Specific implementation mode

[0041] The technical solution of the present invention will be further described in conjunction with the following embodiments.

[0042] Embodiment 1

[0043] A distributed photovoltaic optimization configuration method based on an improved multi - objective cyclic system algorithm, as Figure 2 , is carried out by the following steps:

[0044] S1. Input the input conditions into the CSBO algorithm. The input conditions are low - voltage distribution network area grid parameters, distributed photovoltaic characteristic parameters, and load data. Among them, the low - voltage distribution network area grid parameters are at least one of line resistance, reactance, node voltage level, or transformer parameters; the distributed photovoltaic characteristic parameters are at least one of the power characteristics of photovoltaic cells, conversion efficiency, or the relationship between light intensity and power output; the load data is at least one of the load magnitudes or load curves of each node in the low - voltage distribution network area.

[0045] S2. Generate an initial population. The population consists of multiple individuals, each individual is a Pareto solution, a Pareto solution is a decision variable, and each decision variable includes the distributed photovoltaic access location and rated capacity.

[0046] S3. Simultaneously use fast non - dominated sorting and elite strategy selection to screen the population to obtain the screened population.

[0047] S4. According to the objective value constraints, evolve the screened population in the CSBO algorithm to obtain a new evolved population, and determine whether the iteration termination condition is reached. When it is, define the previous population as the final population and enter S5; when it is not, define the new population as the population and return to S3. The objective values are the voltage fluctuation of the distribution network area obtained through power flow calculation in power system analysis and the line loss rate of the distribution network area obtained through power flow calculation in power system analysis.

[0048] S5. Select a Pareto solution from the corresponding Pareto solution set in the final population, and define the selected Pareto solution as the optimal solution.

[0049] S6. According to the operation constraint conditions, the optimal solution of S5 is cyclically superimposed to obtain the maximum DPV penetration rate, wherein the operation constraint conditions are at least one of the upper and lower voltage limit constraints, the power balance constraint or the line capacity constraint, and the iteration termination condition is the maximum number of iterations.

[0050] It should be noted that the formula used by the CSBO algorithm of the present invention is a power flow calculation. By inputting conditions and the maximum and minimum limits of the target value constraints, the corresponding Pareto solution set in the final population can be obtained. Power flow calculation is a very important analytical calculation for power systems, which is used to study various problems raised in system planning and operation. The use and evolution of the CSBO algorithm is common knowledge for those skilled in the art.

[0051] The multi-objective of the present invention refers to having two target values, namely, the voltage fluctuation of the distribution network area obtained by the flow calculation in the power system analysis and the line loss rate of the distribution network area obtained by the flow calculation in the power system analysis. In the process of solving the CSBO algorithm, the CSBO algorithm continuously adjusts the decision variables (access location and rated capacity) of distributed photovoltaics so that the voltage fluctuation and line loss rate develop in the optimal direction at the same time.

[0052] Since these two objectives often conflict with each other, the present invention is an algorithm for improving the CSBO algorithm. The present invention utilizes a fast non-dominated sorting method based on elite strategy selection to screen out a better individual combination in each iteration, balance the relationship between the two objectives to obtain the final population, and then the management personnel of the substation select the Pareto solution that is biased towards a certain objective function as the optimal solution from the Pareto solution set in the final population. The optimal solution represents a distributed photovoltaic configuration scheme that balances voltage fluctuations and line loss rates to varying degrees.

[0053] In the power system analysis, the voltage fluctuation of the distribution network substation and the line loss rate of the distribution network substation are obtained through the flow calculation. For distribution networks of different levels, the national standards have different fluctuation index limits and line loss rate indicators. The present invention selects voltage fluctuation and line loss rate according to the specific distribution networks of different levels. The target value constraints of the present invention are specifically voltage upper and lower limit constraints, that is, to ensure that the grid voltage fluctuates within the allowable range after connecting to distributed photovoltaics; power balance constraints, that is, to ensure the power balance between distributed photovoltaic power generation and loads and other parts of the power grid; line capacity constraints, that is, to prevent the line from being damaged due to power overload.

[0054] It should also be noted that the low-voltage distribution network parameters of the present invention are used to construct an operation model of the low-voltage distribution network, which is the basis for calculating voltage fluctuations and line loss rates, and affects the operation state of the power grid after the access of distributed photovoltaics. The characteristic parameters of the distributed photovoltaics of the present invention determine the power generation capacity of the distributed photovoltaics under different conditions, and thus affect the power injection situation after their access to the low-voltage distribution network. The load data reflects the electricity consumption demand of the distribution area. When calculating voltage fluctuations and line loss rates, it is necessary to consider the balance relationship between distributed photovoltaic power generation and the load to determine a reasonable access location and capacity. The operation constraint conditions, such as the upper and lower voltage limit constraints, are to ensure that the power grid voltage fluctuates within the allowable range after the access of distributed photovoltaics; the power balance constraint is to ensure the power balance between distributed photovoltaic power generation, the load and other parts of the power grid; the line capacity constraint is to prevent the line from being damaged due to power overload, etc. These operation constraint conditions limit the access schemes of distributed photovoltaics and are the conditions that must be satisfied during the optimization process.

[0055] Among them, S3 is specifically as follows:

[0056] S3.1: Classify the population into different non-dominated fronts through the fast non-dominated sorting method, so as to obtain non-dominated sets at different levels, and enter S3.2;

[0057] S3.2: Select the optimal population from the non-dominated sets at different levels through the elitist strategy to obtain the screened population.

[0058] The specific content of S3.1 is as follows: Define the number of individuals in the previous population as N. Then, the individuals in the previous population are respectively updated after passing through the blood flow model in the CSBO algorithm, passing through the systemic circulation in the CSBO algorithm, and passing through the pulmonary circulation in the CSBO algorithm, and all are merged to obtain a population Rt with 2N individuals. Then, the population Rt is subjected to fast non-dominated sorting to obtain multiple non-dominated sets, and all non-dominated sets are sorted in order from the best to the worst, which are P1, P2,... Pi respectively.

[0059] It should be noted that the fast non-dominated sorting method is a technical means for classifying population individuals into different non-dominated fronts and partitioning Pareto solutions. It can distinguish individuals of different superiority and inferiority degrees in the population, laying a foundation for subsequent evolutionary operations. The elite strategy selection plays a role based on the results of the fast non-dominated sorting. In each iteration, from the Pareto solutions divided by the fast non-dominated sorting, a better individual combination is selected to ensure that these excellent individuals are retained and reproduced during the population evolution process, avoiding the loss of excellent solutions during the evolution process, thereby improving the convergence speed and solution accuracy of the algorithm. The fast non-dominated sorting elite strategy selection and the fast non-dominated sorting method are complementary and closely combined, serving the present invention together. For the fast non-dominated sorting method, it is a conventional and mature sorting method in the art, and those skilled in the art should know its actual sorting method.

[0060] As Figure 3 shown, S3.2 is carried out by the following steps:

[0061] S3.2.1. Directly add the non-dominated set P1 to the new population Qt. When the new population Qt is less than N, enter S3.2.2; when the new population Qt is greater than N, select and add according to the crowding degree comparison method to make the new population Qt equal to N and enter S3.2.4; when the new population Qt is equal to N, enter S3.2.4;

[0062] S3.2.2. Add the subsequent non-dominated sets to the new population Qt in sequence from front to back. After each addition, enter S3.2.3;

[0063] S3.2.3. When the new population Qt is equal to N, stop adding and enter S3.2.4; when the new population Qt is greater than N, select and add according to the crowding degree comparison method to make the new population Qt equal to N and enter S3.2.4; when the new population Qt is less than N, return to S3.2.2;

[0064] S3.2.4. Define the current population Qt as the screened population.

[0065] S6 of the present invention is specifically as follows:

[0066] S6.1. Set the increased value to A, and A > 0 kw, where A can be 1 kw, 5 kw, or 10 kw;

[0067] S6.2. Perform power flow calculation on the superposed increased value in the optimal solution to obtain the superposed value, and enter S6.3;

[0068] S6.3. Judge whether the superposed value exceeds the operation constraint conditions. If not, enter S6.4; if so, enter S6.5;

[0069] S6.4. Let the superposed value be the optimal solution and return to S6.2;

[0070] S6.5. Stop superposition and calculate the maximum DPV permeability based on the current superposition value.

[0071] The present invention can determine the scale and location of distributed photovoltaic DPV in a substation area and is applied to the site selection and configuration settings of distributed photovoltaics; due to the inherent volatility and intermittency of photovoltaic power output, the present invention can improve the impact on the stability of the original distribution network after being connected to the distribution network; the present invention also realizes the reasonable configuration of distributed photovoltaics to maximize the consumption of new energy; at the same time, the present invention helps to accelerate the construction of a new power system with new energy as the main body, while increasing the proportion of new energy installed capacity, ensuring the economic and stable operation of the distribution network.

[0072] The beneficial effects of the distributed photovoltaic optimization configuration method are as follows: First, through the established substation area operation model and photovoltaic system model, considering voltage fluctuations, line loss rates, and various operation constraints, and using the improved multi-objective CSBO algorithm to solve, it can determine the reasonable access location and rated capacity of DPV, effectively reduce the system line loss rate, improve the voltage distribution, and enhance the power supply stability of the low-voltage substation area. Second, the distributed photovoltaic optimization configuration method obtains the distributed photovoltaic access location and rated capacity corresponding to the optimal solution, and can also obtain the maximum photovoltaic permeability that can be accessed in the substation area according to the optimal solution; the present invention provides a key reference for the subsequent investment and construction of photovoltaic power stations, helps to reasonably plan the scale of photovoltaic power stations, and improves the utilization efficiency of new energy. The prior art lacks research on the maximum permeability of DPV that can be configured in the low-voltage substation area, and the present invention makes up for this deficiency. Third, the present invention improves the CSBO algorithm by using the fast non-dominated sorting method based on elite strategy selection. When dealing with multi-objective optimization problems, compared with the single-objective sorting method of the basic CSBO algorithm, it can more reasonably screen the population, more effectively solve the DPV optimization configuration problem in the low-voltage substation area, improve the solution accuracy and efficiency of the algorithm, and has the advantage of strong adaptability of the optimization algorithm.

[0073] Embodiment 2

[0074] A specific application of a distributed photovoltaic optimization configuration method based on an improved multi-objective cyclic system algorithm in Embodiment 1.

[0075] A low-voltage power distribution area in a certain town is selected as the research object. This area contains 10 power consumption nodes, the total length of the line is 5 kilometers, the line resistance is 0.5 Ω / km, the reactance is 0.3 Ω / km, the rated capacity of the transformer is 500 kVA, and the transformation ratio is 10 / 0.4 kV. The average load of each node in the area is 50 kW, the load peak-valley difference is large, the maximum load can reach 80 kW, and the minimum load is 30 kW. The local annual average sunshine hours is 2000 hours. The distributed photovoltaic uses common monocrystalline silicon photovoltaic modules, with a conversion efficiency of 20%, and the power of a single module is 500 W under standard conditions (light intensity 1000 W / m2, cell temperature 25°C).

[0076] Taking the above low-voltage power distribution network parameters, distributed photovoltaic characteristic parameters, and load data as input conditions, input the distributed photovoltaic optimization configuration method based on the improved multi-objective circulation system algorithm in Embodiment 1. Then generate an initial population. Assume that the initial population size is 50 individuals, and each individual represents a combination scheme of DPV access position and rated capacity. Classify the individuals into different non-dominated fronts through fast non-dominated sorting, and divide the Pareto solutions according to the dominance relationship. In the subsequent iterative process, simulate the human circulatory system, perform evolutionary operations on the population, and continuously adjust the access position and rated capacity of the individuals.

[0077] After 200 iterations with a maximum number of iterations, the final population is obtained. The management personnel of the power distribution area select a group of Pareto solutions that comprehensively consider the optimal scheme of voltage fluctuation and line loss rate from the Pareto solution set corresponding to the final population. The optimal solution is to determine that distributed photovoltaics are connected to Node 3 and Node 7. The rated capacity connected to Node 3 is 100 kW, and the rated capacity connected to Node 7 is 80 kW.

[0078] At this time, the voltage fluctuation in the power distribution area is reduced from the original ±8% to within ±3%, and the line loss rate is reduced from 9% to about 5%.

[0079] Finally, in S6, set the added value A to 5 kW, that is, the capacity increased each time is 5 kW. Keep the access positions unchanged at Node 3 and Node 7, and start to cycle and superimpose the capacity. Each time the capacity is increased, it is necessary to check whether the constraints such as the upper and lower limits of the voltage, power balance, and line capacity of the power grid operation are violated.

[0080] After multiple superpositions, it is finally determined that without violating the constraints, the maximum accessible capacity of Node 3 is 150 kW, and the maximum accessible capacity of Node 7 is 120 kW.

[0081] The maximum accessible capacity of Node 3 is 150 kW, and the maximum accessible capacity of Node 7 is 120 kW. Based on these, the maximum PV penetration rate that can be accessed in the substation area is calculated. The calculation method for the maximum PV penetration rate is (150 + 120) / (500 × 0.8) × 100% = 67.5%. 67.5% is the maximum PV penetration rate, where 0.8 is the proportion of the power that the transformer can carry after considering the load factor.

[0082] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the protection scope of the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. A distributed photovoltaic optimization configuration method based on an improved multi-objective cyclic system algorithm, characterized in that It is carried out by the following steps: S1. Input the input conditions into the CSBO algorithm. The input conditions are low-voltage distribution network parameters, distributed photovoltaic characteristic parameters, and load data. S2. Generate an initial population. The population consists of multiple individuals, each individual being a Pareto solution. A Pareto solution is a decision variable, and each decision variable includes the distributed photovoltaic access location and rated capacity. S3. Simultaneously use fast non-dominated sorting and elitist strategy selection to screen the population to obtain the screened population. S4. According to the objective value constraints, evolve the screened population in the CSBO algorithm to obtain a new evolved population. Determine whether the iteration termination condition is reached. When it is, define the previous population as the final population and enter S5; when it is not, define the new population as the population and return to S3. S5. Select a Pareto solution from the corresponding Pareto solution set in the final population, and define the selected Pareto solution as the optimal solution. S6. Perform a cyclic superposition process on the optimal solution in S5 according to the operation constraint conditions to obtain the maximum DPV penetration rate.

2. The distributed photovoltaic optimization configuration method based on the improved multi-objective cyclic system algorithm according to claim 1, wherein The specific content of S3 is as follows: S3.

1. Classify the population into different non-dominated fronts through the fast non-dominated sorting method to obtain non-dominated sets at different levels, and enter S3.

2. S3.

2. Screen out the optimal population from the non-dominated sets at different levels through the elitist strategy selection to obtain the screened population.

3. The distributed photovoltaic optimization configuration method based on the improved multi-objective cyclic system algorithm according to claim 2, wherein The specific content of S3.1 is as follows: Define the number of individuals in the previous population as N. Then, after the individuals in the previous population pass through the blood flow model in the CSBO algorithm, the systemic circulation in the CSBO algorithm, and the pulmonary circulation in the CSBO algorithm respectively and are updated, all the populations are merged to obtain a population Rt with 2N individuals. Then, perform fast non-dominated sorting on the population Rt to obtain multiple non-dominated sets, and sort all the non-dominated sets in order from the best to the worst, which are P1, P2, …… Pi respectively.

4. The distributed photovoltaic optimization configuration method based on the improved multi-objective circulation system algorithm according to claim 3, characterized in that: The specific content of S3.2 is carried out by the following steps: S3.2.

1. Directly add the non-dominated set P1 to the new population Qt. When the new population Qt is less than N, enter S3.2.2; when the new population Qt is greater than N, select and add according to the crowding degree comparison method to make the new population Qt equal to N and enter S3.2.4; when the new population Qt is equal to N, enter S3.2.

4. S3.2.

2. Add the subsequent non-dominated sets to the new population Qt in order from front to back, and enter S3.2.3 after each addition. S3.2.

3. When the new population Qt is equal to N, stop adding and enter S3.2.4; when the new population Qt is greater than N, select and add according to the crowding degree comparison method to make the new population Qt equal to N and enter S3.2.4; when the new population Qt is less than N, return to S3.2.

2. S3.2.

4. Define the current population Qt as the screened population.

5. The distributed photovoltaic optimization configuration method based on the improved multi-objective cyclic system algorithm according to claim 4, wherein: The objective value is the voltage fluctuation of the distribution network substation area obtained through power flow calculation in power system analysis and the line loss rate of the distribution network substation area obtained through power flow calculation in power system analysis.

6. The distributed photovoltaic optimization configuration method based on the improved multi-objective cyclic system algorithm according to claim 5, characterized in that, The specific content of S6 is as follows: S6.

1. Set the incremental value as A, and A > 0 kw; S6.

2. Conduct a power flow calculation on the superimposed incremental value in the optimal solution to obtain the superimposed value, and proceed to S6.3; S6.

3. Determine whether the superimposed value exceeds the operating constraint conditions. If not, proceed to S6.4; if so, proceed to S6.5; S6.

4. Set the superimposed value as the optimal solution and return to S6.2; S6.

5. Stop superimposing, and calculate the maximum DPV penetration rate based on the current superimposed value.

7. The distributed photovoltaic optimization configuration method based on the improved multi-objective cyclic system algorithm according to claim 6, characterized in that: The low-voltage substation grid parameters are at least one of line resistance, reactance, node voltage level, or transformer parameters.

8. The distributed photovoltaic optimization configuration method based on the improved multi-objective cyclic system algorithm according to claim 7, characterized in that: The distributed PV characteristic parameters are at least one of the power characteristics of PV cells, conversion efficiency, or the relationship between light intensity and power output.

9. The distributed photovoltaic optimization configuration method based on the improved multi-objective cyclic system algorithm according to claim 8, characterized in that: The load data is at least one of the load magnitudes or load curves of each node in the low-voltage substation.

10. The distributed photovoltaic optimization configuration method based on the improved multi-objective cyclic system algorithm according to claim 9, characterized in that: The operating constraint conditions are at least one of voltage upper and lower limit constraints, power balance constraints, or line capacity constraints; The iteration termination condition is the maximum number of iterations.

Images