Optimization method and system for photovoltaic access distribution network

Abstract

The application discloses an optimization method and system for photovoltaic access distribution network, and the method comprises the following steps: based on the photovoltaic construction, a certain number of populations are determined; through Newton-Raphson power flow calculation, the distribution network operation parameters corresponding to each gene individual are obtained; based on the distribution network operation parameters, a multi-objective function and a constraint condition, the gene individuals are screened; the screened gene individuals are iteratively calculated to obtain optimal offspring of each population; the optimal offspring is introduced into an elite population; if a set iteration end criterion is met, an optimal individual gene representative photovoltaic access scheme is obtained; otherwise, the above operation is performed again; based on the distribution network operation parameters and a pre-established multi-objective function and constraint condition, the gene individuals are screened, the defect of immature convergence is overcome, and the global optimization can be achieved in the photovoltaic access optimization process; the screened gene individuals are iteratively calculated to obtain optimal offspring of each population, and the local optimization is performed, so that the balanced global search and local search performance are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of electric power new energy optimization planning, and in particular to an optimization method and system for photovoltaic access to a distribution network. Background Art

[0002] In the era of sustainable energy development, distributed generation technologies based primarily on renewable energy have become an effective supplement to large power grids due to their flexible, reliable, clean, and environmentally friendly nature, and have experienced rapid growth. Distributed photovoltaic power generation, as one of the most effective ways to utilize new energy, primarily relies on integration into distribution networks. However, this integration alters the topology and power flow distribution of the existing distribution system. Furthermore, the output characteristics of distributed photovoltaics are intermittent and fluctuating due to environmental factors, thus impacting the safe and stable operation of the power grid. With the integration of large numbers of high-penetration photovoltaics, it is crucial to study optimization methods for integrating distributed photovoltaics into distribution networks to ensure safe and stable power grid operation.

[0003] Regarding optimization models for distributed photovoltaic access to distribution networks, some literature considers the impact of photovoltaic access on a specific power quality factor in the distribution network and establishes single-objective optimization models for optimization planning, such as mathematical models for limiting distribution network node voltage. These single-objective optimization models only consider a specific power quality indicator, and the planning results cannot guarantee the safe and stable operation of the distribution network under other indicators, making them not universally applicable. Other literature considers photovoltaic access planning from the perspective of planning economics, establishing mathematical models with economic efficiency as the objective to pursue economically optimal planning solutions. However, these models must be based on stable system operation to ensure meaningful planning results. Combining the above two types of optimization planning models, some literature also establishes a two-tiered planning model for photovoltaic access. The lower-tier model ensures stable distribution network operation, while the upper-tier planning focuses on the lower-tier optimal solution while also considering distribution network economics. The advantage of hierarchical planning is that it reduces computational complexity and avoids dealing with multiple objective functions. However, when there are conflicts between the upper and lower-tier planning objectives, the lower-tier model cannot fully consider the adverse impact of the upper-tier optimization objective on the lower-tier optimization decision, and the resulting solution cannot guarantee the optimal solution for the system.

[0004] Methods for solving photovoltaic access optimization models include traditional algorithms and intelligent algorithms. However, since the general optimization planning model is a nonlinear problem with multiple constraints, it is complex and large in scale. At the same time, the constraints make the feasible search space of decision variables irregular, making it difficult for traditional gradient-based algorithms to effectively find the optimal solution. Therefore, related research often uses intelligent algorithms such as genetic algorithms and particle swarm optimization. However, the optimization performance of these algorithms is greatly affected by the optimization parameters. Improper parameter settings can lead to premature convergence to local optimal solutions. The optimal parameter setting requires multiple experiments to select the optimal parameters. In summary, how to effectively solve photovoltaic access optimization models is also a key issue in this type of research. Summary of the Invention

[0005] In order to solve the problem of lack of effective solution for photovoltaic access optimization method in the prior art, the present invention provides an optimization method and system for photovoltaic access to distribution network.

[0006] The technical solution provided by the present invention is:

[0007] A method for optimizing photovoltaic access to a distribution network, comprising:

[0008] Step 1: A set number of populations are constructed based on multiple photovoltaic cells to be connected to the distribution network. The distribution network operation parameters corresponding to each gene individual in each population are obtained through Newton-Raphson power flow calculation.

[0009] Step 2: Screening the gene individuals based on the distribution network operation parameters and pre-established multi-objective functions and constraints;

[0010] Step 3: Iteratively calculate the gene individuals obtained by screening to obtain the optimal offspring of each population;

[0011] Step 4: Introduce the optimal offspring into the elite population. If the set iteration end criterion is met, the photovoltaic access plan represented by the optimal individual gene is obtained; otherwise, execute step 1 again;

[0012] The distribution network operating parameters include: distribution network node voltage, line current and distributed photovoltaic capacity.

[0013] Preferably, the photovoltaic power generation system to be connected to the power distribution network is used to construct a set number of populations, and the power distribution network operation parameters corresponding to each gene individual in each population are obtained through Newton-Raphson power flow calculation, including:

[0014] Based on the photovoltaics to be connected to the distribution network, a set number of photovoltaics are used to construct a genetic individual, and the set number of genetic individuals are used to construct a population;

[0015] The access capacity of each gene individual is expressed through the decimal code of the photovoltaic genome;

[0016] The gene individuals are normalized and sequentially brought into Newton-Raphson power flow calculation according to the set access positions to obtain the distribution network operation parameters corresponding to each gene individual.

[0017] Preferably, the establishment of the objective function and constraint conditions includes:

[0018] The objective function is constructed by maximizing the sum of grid-connected photovoltaic capacity, minimizing operation and maintenance costs, and minimizing grid-connected active power losses;

[0019] Constraint conditions are constructed based on voltage deviation constraints, line current constraints, voltage fluctuation constraints, harmonic size constraints and power flow balance relationship constraints.

[0020] Preferably, the screening of the gene individuals based on the distribution network operation parameters and pre-established multi-objective functions and constraints includes:

[0021] Based on each population, determine whether the distribution network node voltage and line current corresponding to each gene individual in the current population meet the voltage deviation constraint and the line current carrying capacity constraint; determine whether the distributed photovoltaic capacity in the current population meets the harmonic size constraint and the voltage fluctuation constraint; if all are satisfied, the fitness of the gene individual is non-zero; otherwise, the fitness of the gene individual is zero;

[0022] The gene individuals are screened based on the fitness. If the fitness of all gene individuals in the current population is non-zero, the gene individuals in the current population are introduced into the adjacent population according to the set generation number, and the gene individuals with non-zero results in various populations form a mating pool; otherwise, they are not introduced into the adjacent population.

[0023] Preferably, the fitness is calculated by the following formula:

[0024]

[0025] Among them, F is the calibration fitness of the k-generation gene individual; f is the calculated value of the objective function of the k-generation gene individual; is the minimum objective function value among the k-generation genetic individuals; ξ k is the k-generation pressure adjustment value, which decreases as k increases.

[0026] Preferably, the iterative calculation of the gene individuals obtained by screening to obtain the optimal offspring of each population includes:

[0027] Step 3-1: Based on the fitness of the genetic individuals, randomly select the father and mother of the genetic individuals in the mating pool through a roulette wheel method, and obtain a random number;

[0028] Step 3-2: If the random number is less than the set crossover probability, the parent and parent generations are crossovered to obtain the offspring genome; otherwise, no crossover is performed and the parent genome is used as the offspring genome;

[0029] Step 3-3: If the random number is less than the set mutation probability, a mutation operation is performed on the offspring genome; otherwise, no mutation operation is performed;

[0030] Step 3-4: Re-execute step 3-1 until each population generates a progeny genome, and select the optimal progeny from each population.

[0031] A photovoltaic access distribution network optimization system, comprising:

[0032] The distribution network operation parameter acquisition module is used to construct a set number of populations based on multiple photovoltaics to be connected to the distribution network, and obtain the distribution network operation parameters corresponding to each gene individual in each population through Newton-Raphson power flow calculation;

[0033] A screening module, configured to screen the gene individuals based on the distribution network operation parameters and pre-established multi-objective functions and constraints;

[0034] The optimal offspring screening module is used to iteratively calculate the optimal offspring of each population based on the gene individuals screened;

[0035] An access scheme confirmation module is used to introduce the optimal offspring into the elite population. If the set iteration end criterion is met, the photovoltaic access scheme represented by the optimal individual gene is obtained; otherwise, step 1 is executed again;

[0036] The distribution network operation parameters obtained in the distribution network operation parameter acquisition module include: distribution network node voltage, line current and distributed photovoltaic capacity.

[0037] Preferably, the distribution network operation parameter acquisition module includes:

[0038] Gene individual construction submodule: based on the photovoltaic power plants to be connected to the distribution network, a gene individual is constructed with a set number of photovoltaic power plants, and a population is constructed with the set number of gene individuals;

[0039] Access capacity construction submodule: The access capacity of each gene individual is expressed through the decimal code of the photovoltaic genome;

[0040] The power flow calculation submodule: normalizes the gene individuals and sequentially brings them into the Newton-Raphson power flow calculation according to the set access positions to obtain the distribution network operation parameters corresponding to each gene individual.

[0041] Preferably, the screening module includes:

[0042] Objective function construction submodule: The objective function is constructed by maximizing the sum of grid-connected photovoltaic capacity, minimizing operation and maintenance costs, and minimizing grid-connected active power loss;

[0043] Constraint construction submodule: Constructs constraints based on voltage deviation constraints, line current constraints, voltage fluctuation constraints, harmonic size constraints and power flow balance relationship constraints.

[0044] Preferably, the screening module further includes:

[0045] Fitness calculation submodule: Based on each population, it determines whether the distribution network node voltage and line current corresponding to each gene individual in the current population meet the voltage deviation constraint and line current carrying capacity constraint; it determines whether the distributed photovoltaic capacity in the current population meets the harmonic size constraint and voltage fluctuation constraint; if all are satisfied, the fitness of the gene individual is non-zero; otherwise, the fitness of the gene individual is zero;

[0046] Screening submodule: Screen the gene individuals based on the fitness. If the fitness of all gene individuals in the current population is non-zero, the gene individuals in the current population are introduced into the adjacent population according to the set generation number, and the gene individuals with non-zero results in various populations form a mating pool; otherwise, they are not introduced into the adjacent population.

[0047] Preferably, the fitness calculation submodule calculates the fitness using the following formula:

[0048]

[0049] Among them, F is the calibration fitness of the k-generation gene individual; f is the calculated value of the objective function of the k-generation gene individual; is the minimum objective function value among the k-generation genetic individuals; ξ k is the k-generation pressure adjustment value, which decreases as k increases.

[0050] Preferably, the optimal offspring screening module includes:

[0051] Father and mother screening submodule: based on the fitness of the gene individuals, randomly select the father and mother from the gene individuals in the mating pool through the roulette method and obtain random numbers;

[0052] Crossover operation submodule: If the random number is less than the set crossover probability, the parent and parent generations are crossovered to obtain the offspring genome; otherwise, no crossover operation is performed and the parent genome is used as the offspring genome;

[0053] Mutation operator module: if the random number is less than the set mutation probability, the offspring genome is subjected to mutation operation; otherwise, no mutation operation is performed;

[0054] Iteration submodule: Rerun the parent and parent screening submodule until each population generates offspring genomes and selects the best offspring from each population.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] The technical solution provided by the present invention includes: constructing a set number of populations based on multiple photovoltaic systems to be connected to the distribution network; obtaining the distribution network operating parameters corresponding to each gene individual in each population through Newton-Raphson power flow calculation; screening the gene individuals based on the distribution network operating parameters and pre-established multi-objective functions and constraints; iteratively calculating the screened gene individuals to obtain the optimal offspring of each population; introducing the optimal offspring into an elite population; if the set iteration end criteria are met, a photovoltaic access solution represented by the optimal individual gene is obtained; otherwise, the above operations are performed again; the distribution network operating parameters include: distribution network node voltage, line current, and distributed photovoltaic capacity. This solution overcomes the drawback of premature convergence by screening the gene individuals based on the distribution network operating parameters and pre-established multi-objective functions and constraints, enabling global optimization during the photovoltaic access optimization process; iteratively calculating the screened gene individuals to obtain the optimal offspring of each population, and performing local optimization, thereby achieving the goal of balancing global and local search performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a flow chart of a method for optimizing photovoltaic access to a distribution network according to the present invention;

[0058] Figure 2 The photovoltaic grid-connected optimization modeling and solution flow chart of the present invention;

[0059] Figure 3 This is a topology diagram of the IEEE33 node distribution network according to an embodiment of the present invention. DETAILED DESCRIPTION

[0060] In order to better understand the present invention, the present invention is further described below with reference to the accompanying drawings and examples.

[0061] Example 1:

[0062] This embodiment provides a method for optimizing photovoltaic access to a distribution network. The method flow chart is as follows: Figure 1 shown.

[0063] Step 1:

[0064] The basic parameters of the distribution network system to which the photovoltaic system is connected are considered to facilitate the subsequent model establishment and solution, such as system voltage level, nominal voltage, benchmark short-circuit capacity, distribution network structure topology, number of nodes, number of branches, and the presence or absence of tie lines.

[0065] Step 2:

[0066] Considering the access planning of multiple photovoltaic systems, in order to improve the utilization rate of photovoltaic systems, the maximum sum of the capacity of multiple photovoltaic systems connected to the grid is the first economic sub-goal.

[0067]

[0068] Where, F1 is the total grid capacity function; S DGi is the access capacity of the i-th photovoltaic unit; N is the number of photovoltaic installed grid-connected units.

[0069] Step 3:

[0070] In order to improve the economic efficiency of the distribution network, the second sub-goal is to minimize the investment, operation and maintenance costs of distributed photovoltaics.

[0071]

[0072]

[0073]

[0074] In the formula, Y1 is the investment cost (10,000 yuan); K DG is the investment cost of photovoltaic power per unit capacity (10,000 yuan / kVA); Y2 is the operation and maintenance cost (10,000 yuan); t is the service life of photovoltaic power (years); r is the discount rate; T DG is the photovoltaic operation and maintenance cost per unit capacity (10,000 yuan / kVA); Y3 is the photovoltaic power generation benefit; E i is the annual power generation of each photovoltaic unit (kW·h); γ is the power generation efficiency (10,000 yuan / kW·h).

[0075] Then the sub-objective function for minimizing the investment operation and maintenance cost is:

[0076] minF2=Y1+Y2-Y3

[0077] Where, F2 is the investment operation and maintenance cost function.

[0078] Step 4:

[0079] In order to improve the economic efficiency of the distribution network, the third sub-goal is to minimize the active power loss of the distribution network after the distributed photovoltaic is connected.

[0080]

[0081] Where, F3 is the active power loss function of the distribution network; M is the number of branches in the distribution network system; G h,ij is the conductance of branch h, i, j are the node numbers on both sides of the branch, U h,i ,U h,jis the effective value of the voltage at node i, j, δ ij is the phase difference between the voltages at nodes i and j.

[0082] Step 5:

[0083] The three economic objective functions are combined into the overall economic objective function using the linear weighted combination method.

[0084] maxF=ω1F1-ω2F2-ω3F3

[0085] Where F is the economic objective function; ω1, ω2, ω3 are the weights of the three sub-objective functions.

[0086] Step 6:

[0087] Considering the voltage deviation constraint, determine planning constraint one. According to GB / T12325-2008 "Power Quality Supply Voltage Deviation", lines of different voltage levels have different allowable voltage deviation limits. Therefore, the voltage deviation constraint is:

[0088] U N (1-ε)≤U i ≤U N (1+ε)

[0089] Where U N is the system nominal voltage, and ε is the voltage deviation rate limit specified by the system voltage level.

[0090] Step 7:

[0091] Considering the line current constraints, determine the second planning constraint. To prevent the line current from being too large and causing overload, the line current is constrained:

[0092] |I k |≤Ik ,max

[0093] Where, I k is the current value on line k; I k,max is the maximum current value allowed to be delivered by line k.

[0094] Step 8:

[0095] Considering the voltage fluctuation constraint, determine the third planning constraint. According to GB / T12325-2008 "Power Quality - Voltage Fluctuation and Flicker", the allowable voltage fluctuation limits for systems with different voltage levels and voltage fluctuation frequencies are different. The voltage fluctuation constraint is determined as follows:

[0096] d k %≤d max %

[0097] Where, d k % is the system voltage fluctuation value caused by the photovoltaic system, d max % is the maximum voltage fluctuation value specified by the national standard.

[0098] GB / T12325-2008 "Power Quality - Voltage Fluctuations and Flicker" provides a method for estimating voltage fluctuations. When the changes in active power and reactive power ΔP and ΔQ of the three-phase load are known, the following formula can be used for estimation:

[0099]

[0100] Where R L ,X L are line resistance and reactance, ΔP and ΔQ are load active and reactive changes. Considering the load is a constant power load, the active and reactive changes here are the active and reactive changes of the photovoltaic output.

[0101] Step 9:

[0102] Considering the harmonic size constraints, determine the fourth planning constraint. According to GB / T14549-1993 "Power Quality Public Grid Harmonics", systems of different voltage levels have clear limits on the allowable values ​​of each harmonic current injected into the access point. At the same time, the odd harmonic voltage content rate, even harmonic voltage content rate and voltage total harmonic distortion rate are also specified. The critical values ​​that can be reached are also specified.

[0103] To determine the constraints imposed by harmonic currents on photovoltaic access, according to GB / T14549-1993, "Harmonics in Public Grids for Power Quality," the maximum allowable harmonic currents for injection are specified for systems with different voltage levels and reference short-circuit capacities. Furthermore, according to GB / T19939-2005, "Technical Requirements for Photovoltaic System Grid Connection," the maximum allowable current harmonic content of photovoltaic grid-connected power generation systems is specified. To ensure the effectiveness of the harmonic current constraint, the injected current of the photovoltaic system at maximum harmonic output is used to determine its maximum access capacity. The constraint determined by the harmonic current is:

[0104]

[0105] Where S DG is the access photovoltaic capacity; U N is the system nominal voltage; μ k The harmonic content rate limit of each output of the photovoltaic system specified in GB / T19939-2005 "Technical Requirements for Grid-Connected Photovoltaic Systems", I k These are the harmonic currents allowed to be connected to the power grid as specified in GB / T14549-1993 "Power Quality Harmonics in Public Grids", where the subscript k is the harmonic order.

[0106] This study examines the constraints imposed by harmonic voltage content and total harmonic distortion (THH) on PV integration. GB / T14549-1993, "Power Quality - Harmonics in Public Grids," specifies the permissible harmonic voltage content and THD for systems of different voltage levels. To ensure the validity of the harmonic voltage content constraint, the PV system's injected current at maximum harmonic output is used to determine its access capacity.

[0107]

[0108]

[0109] U h =HRU h ×U N

[0110]

[0111] Where I k is the harmonic current injected by photovoltaic; k is the harmonic order; HRU k is the voltage content rate of each harmonic; S k is the system short-circuit capacity; U h is the harmonic voltage; T HDU is the total harmonic voltage distortion rate. The constraint conditions determined by the odd and even harmonic voltage content rates and the voltage total harmonic distortion rate are:

[0112] HRU k ≤0.032k=2n+1,n=1,2,3…

[0113] HRU k ≤0.016k=2n,n=1,2,3…

[0114] T HDU ≤0.04

[0115] Connected photovoltaic capacity S DG It should also meet the requirements of the harmonic current value limit, odd harmonic voltage content rate limit, even harmonic voltage content rate limit, and voltage total harmonic distortion rate limit of the distribution network system after access.

[0116] Step 10:

[0117] Considering the constraints of the power flow balance, determine the fifth planning constraint:

[0118]

[0119] Where, P DG,k ,Q DG,k is the active and reactive power output of the photovoltaic connected at node k; P k ,Qk is the active and reactive load of k nodes; G kj ,B kj are the conductance and susceptance in the node admittance matrix.

[0120] Step 11:

[0121] Combining the overall economic objective function and power quality constraints, a photovoltaic grid-connected optimization model that takes both economy and safety into consideration is determined:

[0122]

[0123] In the formula, the objective function is composed of the photovoltaic capacity S DG , voltage U, phase δ, line conductance G are determined, S DG Through the power flow equation and active and reactive load P, Q, system voltage U, current I, phase δ line conductance G, network loss P 损 Establish a connection, the g function is the voltage deviation and line thermal constraint determined by voltage U and current I, and the m function is the capacity S DG Determined voltage fluctuation and harmonic constraints.

[0124] Step 12:

[0125] Determine the distribution network system structure parameters and the number and location of photovoltaic access points. Determine the topology of the distribution network and its network parameters in the photovoltaic planning problem, including line impedance, active and reactive load size. Determine the number and location of photovoltaic access points to be planned for subsequent optimization planning. PV grid optimization modeling and solution flow chart, such as Figure 2 shown.

[0126] Step 13:

[0127] Initialize the code and determine the genome of each initial genetic population.

[0128] Constrained by distributed photovoltaic power output 0<S PV <aMW and photovoltaic capacity discrete accuracy ξ=bKW, the binary code length of the individual genome of a single photovoltaic capacity is determined as:

[0129]

[0130] Since we are studying the optimization of multiple photovoltaic access, each individual in the population contains N photovoltaic individuals. The genotype code length of each individual is LN. The number of individuals in the population is 100, and the population size is 10. Therefore, the initialization uses a random function to generate 1000 individual genotypes with a code length of LN. This is the genetic initialization of each individual in each initial population.

[0131] Step 14:

[0132] The Newton-Raphson power flow algorithm is used to solve the operating parameters of the active distribution network.

[0133] For each individual in each population, its binary genotype is converted into a decimal phenotype, representing the N photovoltaic access capacities it represents. This phenotype is then normalized and entered into a Newton-Raphson power flow calculation program based on the access location. The distribution network power flow distribution, line currents, and node voltages at the corresponding access location are calculated. This power flow calculation program is repeated to determine the operating parameters of the distribution network system under the access scenarios represented by all individuals in each population.

[0134] Step 15:

[0135] Fitness calculation determines whether the access scheme represented by each individual in each group meets the various power quality constraints, whether the voltage and line current of each node in the flow calculation meet the voltage deviation and line current carrying capacity constraints, and whether the distributed photovoltaic capacity meets the harmonic and voltage fluctuation constraints. If a constraint is not met, the fitness of the individual is set to f(k) = 0. If all the above power quality constraints are met, the value of the fitness function is based on the size of the economic objective function. The size of the value reflects the economic cost of the access scheme. The larger the value, the better the economic performance. However, considering that the photovoltaic capacity and network active power loss obtained by the flow calculation are all in per-unit system, the final fitness value expressed is very small, the difference between different individuals is small, and the probability of being selected is very close, which affects the effectiveness of the selection function. Therefore, this paper uses the dynamic linear calibration method to calibrate the fitness function to improve the effectiveness of the selection action. The fitness function of the inherited k-generation individual is:

[0136]

[0137] Where F is the value of the k-generation individual calibration fitness function; f is the calculated value of the k-generation individual economic objective function; is the minimum objective function value among k generations of individuals; ξ k k represents the pressure adjustment value, which decreases as k increases. The present invention adopts the following setting method:

[0138]

[0139] Where M is the initial value of the pressure adjustment value; c is the reduction coefficient of the pressure adjustment value, and its value range is [0.9, 0.999].

[0140] At the same time, to achieve multi-population collaborative optimization, after a certain number of iterations, the immigration operator is used to achieve communication between multiple populations, and the individuals with the best fitness in the k population are introduced into the k+1 population. Finally, the individuals with non-zero fitness function values ​​in a population form the mating pool of that population.

[0141] The fitness level is determined according to the economic objective calibrated by the fitness function. Individuals with non-zero fitness form the mating pool of that population. If all conditions are met, the fitness value is set according to the fitness function calibrated by dynamic linear calibration. The individuals with the best fitness from the k population are introduced into the k+1 population over a certain number of generations. Finally, individuals with non-zero fitness from each population form the mating pool.

[0142] Step 16:

[0143] To perform genetic operations, first, individuals in the mating pool form a roulette wheel according to their fitness. The greater the fitness, the greater the probability of the individual being selected. The roulette wheel is used to randomly select the father and mother. The probability of different individuals being selected is:

[0144]

[0145] Where, F i is the fitness function value of the i-th individual in the mating pool, P i is the probability of the i-th individual being selected in the mating pool.

[0146] After randomly selecting a parent or parent using a roulette wheel, a crossover is performed. A random function is first used to generate a random number. When the random number is less than the crossover probability, a crossover is performed, using the random function to determine the location where the parent and parent genomes intersect. The parental genes are then crossed to generate a new individual genome. Conversely, when the random number is greater than the crossover probability, no crossover is performed, and the parental genes are selected as the genome of the new offspring. To maximize optimization performance and speed due to the large number of populations and individuals, an adaptive crossover probability is introduced, which changes as the fitness function changes.

[0147]

[0148] Where, P c is the adaptive mutation probability; f max is the maximum fitness value in each generation of the population; f avg is the average fitness value of all individuals in the population; f is the larger fitness value of the two individuals to be crossed; k1 and k3 are proportional constants.

[0149] The advantage of introducing adaptive crossover probability is that a smaller crossover probability is set for individuals with fitness higher than the average value so that the excellent individuals of the population can be retained, and individuals with fitness lower than the average value are set to the constant crossover probability set for the population for optimization, which can effectively improve the performance of each group in optimizing its own objective function.

[0150] After obtaining a new offspring through crossover, a mutation operation is performed. A random function is used to generate a random number. When the random value is greater than the mutation probability, no mutation is performed on the new offspring. However, when the random value is less than the mutation probability, a mutation operation is performed on the new offspring at a random mutation position to expand the diversity of individuals in the population. At the same time, due to the large number of populations and individuals, an adaptive mutation probability is introduced to maximize optimization performance and speed. The mutation probability changes as the fitness function changes.

[0151]

[0152] Where, P m is the adaptive mutation probability; f max is the maximum fitness value in each generation of the population; f avg is the average fitness value of all individuals in the population; f' is the fitness value of the individual for determining whether to perform mutation action; k3, k4 are proportional constants.

[0153] Similarly, the advantage of introducing adaptive crossover probability is that for individuals with fitness higher than the average, a smaller mutation probability is set so that the excellent individuals of the population can be retained, and for individuals with fitness lower than the average, the constant mutation probability set for the population is set for optimization, which can effectively improve the performance of each group in optimizing its own objective function.

[0154] Repeat the above selection, crossover, and mutation actions until each population generates a new generation of offspring.

[0155] Step 17:

[0156] After generating new offspring, the best offspring of all individuals in each population are introduced into the elite population to determine whether the best individual in the elite population meets the iteration end criterion. If it does, the photovoltaic access scheme represented by the genome of the best individual is output. If not, go to step 13 and repeat the iterative optimization until the iteration end criterion is met.

[0157] Example 2:

[0158] Step 1: Determine the basic parameters of the distribution network system. Select the IEEE33 node distribution network model, whose structure topology is as follows: Figure 3 The system consists of 4 feeders with a total of 33 nodes. The dotted line is the system tie line. The system voltage level is 10KV. Node 1 represents the low-voltage side of the substation and is set as a balance node with a voltage value of 1.05pu. The other nodes are PQ nodes. The total active load is 3715kW, the total reactive load is 2300kvar, and the line voltage reference value is U B =10KV, three-phase power reference value S B =10000KV·A, and study the photovoltaic planning when the access locations are 7, 9, 11, and 13.

[0159] Step 2: Considering the access capacity, determine the economic objective function 1:

[0160]

[0161] Step 3: Considering the investment, operation and maintenance costs of photovoltaics, determine the second economic objective function as follows:

[0162]

[0163] Step 4: The third sub-goal is to minimize the active power loss of the distribution network after the distributed photovoltaic is connected.

[0164]

[0165] Step 5: Use the linear weighted combination method to combine the three economic objective functions into the overall economic objective function;

[0166] maxF=0.6ω1F1-0.2F2-0.2F3

[0167] Step 6: Considering the voltage deviation constraint and the system voltage level of 10 kV, determine planning constraint condition 1;

[0168] 0.93U N ≤U i ≤1.07U N

[0169] Step 7: Consider the line current constraints and determine the second planning constraint;

[0170] |I k |≤175A

[0171] Step 8: Consider the voltage fluctuation constraint and determine the third planning constraint;

[0172] d k %≤3%

[0173] Step 9: Considering the harmonic size constraint, determine the fourth planning constraint;

[0174]

[0175] HRU k ≤0.032 k=2n+1,n=1,2,3…

[0176] HRU k ≤0.016 k=2n,n=1,2,3…

[0177] T HDU ≤0.04

[0178] The harmonic limits for the 10KV system are shown in the following table.

[0179] Table 1 Allowable values ​​of harmonic currents

[0180]

[0181] Step 10: Consider the power flow balance constraint and determine the fifth planning constraint;

[0182]

[0183] Step 11: Determine the photovoltaic grid-connected optimization model based on the overall economic objective function and power quality constraints

[0184] Step 12: Determine the distribution network system structure parameters and the number and location of photovoltaic access points. Use the Newton-Raphson method to write a power flow calculation program to calculate the system node voltage, line current, power flow distribution, etc. when photovoltaic access is applied. The impedance and load size of some distribution network lines are shown in the following table.

[0185] Table 2 Impedance of distribution network lines and node loads

[0186]

[0187] Step 13: Initialize the code and determine the genome of each initial genetic population, and the distributed photovoltaic power output is constrained to 0<S PV <6MW and the discrete precision of photovoltaic capacity ξ=1KW, the binary code length of the photovoltaic capacity genome is determined as:

[0188]

[0189] The genomes of the initial individuals in various populations are generated using random functions, and the code length is 54.

[0190] Step 15: Fitness calculation,

[0191] The fitness function of the k-generation individuals is and the pressure adjustment value is set as:

[0192]

[0193]

[0194] Step 16: Perform genetic operations, use the roulette wheel method to select parents from the mating pool, and use the adaptive crossover and mutation probabilities to perform crossover and mutation operations. The adaptive crossover and mutation probabilities are as follows:

[0195]

[0196]

[0197] Wherein, k2 and k4 are set differently in different populations to achieve the emphasis on different optimization performances. A larger value is biased towards global optimization, while a smaller value focuses on local optimization.

[0198] Step 17: After genetic optimization of multiple populations, select the optimal solution from each population and introduce it into the elite population. Determine whether the elite individuals introduced in this generation are the same as the elite individuals of the previous generation. If they remain the same for five generations, the optimization is considered to be stable and the access plan represented by the optimal solution is output. Otherwise, go to step 13 and repeat the optimization iteration.

[0199] The planned capacity of the photovoltaic access locations at 7, 9, 11, and 13 o'clock, finally obtained through optimization, is:

[0200] PV1=303kW, PV2=695kW, PV3=852kW, PV4=803kW,

[0201] Example 3:

[0202] This embodiment provides an optimization system for photovoltaic access to a distribution network, including:

[0203] The distribution network operation parameter acquisition module is used to construct a set number of populations based on multiple photovoltaics to be connected to the distribution network, and obtain the distribution network operation parameters corresponding to each gene individual in each population through Newton-Raphson power flow calculation;

[0204] A screening module, configured to screen the gene individuals based on the distribution network operation parameters and pre-established multi-objective functions and constraints;

[0205] The optimal offspring screening module is used to iteratively calculate the optimal offspring of each population based on the gene individuals screened;

[0206] An access scheme confirmation module is used to introduce the optimal offspring into the elite population. If the set iteration end criterion is met, the photovoltaic access scheme represented by the optimal individual gene is obtained; otherwise, step 1 is executed again;

[0207] The distribution network operation parameters obtained in the distribution network operation parameter acquisition module include: distribution network node voltage, line current and distributed photovoltaic capacity.

[0208] The distribution network operation parameter acquisition module includes:

[0209] Gene individual construction submodule: based on the photovoltaic power plants to be connected to the distribution network, a gene individual is constructed with a set number of photovoltaic power plants, and a population is constructed with the set number of gene individuals;

[0210] Access capacity construction submodule: The access capacity of each gene individual is expressed through the decimal code of the photovoltaic genome;

[0211] The power flow calculation submodule: normalizes the gene individuals and sequentially brings them into the Newton-Raphson power flow calculation according to the set access positions to obtain the distribution network operation parameters corresponding to each gene individual.

[0212] The screening module includes:

[0213] Objective function construction submodule: The objective function is constructed by maximizing the sum of grid-connected photovoltaic capacity, minimizing operation and maintenance costs, and minimizing grid-connected active power loss;

[0214] Constraint construction submodule: Constructs constraints based on voltage deviation constraints, line current constraints, voltage fluctuation constraints, harmonic size constraints and power flow balance relationship constraints.

[0215] The screening module further includes:

[0216] Fitness calculation submodule: Based on each population, it determines whether the distribution network node voltage and line current corresponding to each gene individual in the current population meet the voltage deviation constraint and line current carrying capacity constraint; it determines whether the distributed photovoltaic capacity in the current population meets the harmonic size constraint and voltage fluctuation constraint; if all are satisfied, the fitness of the gene individual is non-zero; otherwise, the fitness of the gene individual is zero;

[0217] Screening submodule: Screen the gene individuals based on the fitness. If the fitness of all gene individuals in the current population is non-zero, the gene individuals in the current population are introduced into the adjacent population according to the set generation number, and the gene individuals with non-zero results in various populations form a mating pool; otherwise, they are not introduced into the adjacent population.

[0218] The fitness calculation submodule calculates the fitness using the following formula:

[0219]

[0220] Among them, F is the calibration fitness of the k-generation gene individual; f is the calculated value of the objective function of the k-generation gene individual; is the minimum objective function value among the k-generation genetic individuals; ξ k is the k-generation pressure adjustment value, which decreases as k increases.

[0221] The optimal offspring screening module includes:

[0222] Father and mother screening submodule: based on the fitness of the gene individuals, randomly select the father and mother from the gene individuals in the mating pool through the roulette method and obtain random numbers;

[0223] Crossover operation submodule: If the random number is less than the set crossover probability, the parent and parent generations are crossovered to obtain the offspring genome; otherwise, no crossover operation is performed and the parent genome is used as the offspring genome;

[0224] Mutation operator module: if the random number is less than the set mutation probability, the offspring genome is subjected to mutation operation; otherwise, no mutation operation is performed;

[0225] Iteration submodule: Rerun the parent and parent screening submodule until each population generates offspring genomes and selects the best offspring from each population.

[0226] Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0227] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0228] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0229] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0230] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0231] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention are included in the scope of the claims of the present invention to be approved.

Claims

1. A method for optimizing photovoltaic access to a distribution network, characterized in that: include: Step 1: A set number of populations are constructed based on multiple photovoltaic cells to be connected to the distribution network. The distribution network operation parameters corresponding to each gene individual in each population are obtained through Newton-Raphson power flow calculation. Step 2: Screening the gene individuals based on the distribution network operation parameters and pre-established multi-objective functions and constraints; Step 3: Iteratively calculate the gene individuals obtained by screening to obtain the optimal offspring of each population; Step 4: Introduce the optimal offspring into the elite population. If the set iteration end criterion is met, the photovoltaic access plan represented by the optimal individual gene is obtained; Otherwise, execute step 1 again; The distribution network operating parameters include: distribution network node voltage, line current and distributed photovoltaic capacity; The establishment of the objective function and constraint conditions includes: The objective function is constructed by maximizing the sum of grid-connected photovoltaic capacity, minimizing operation and maintenance costs, and minimizing grid-connected active power losses; Construct constraint conditions based on voltage deviation constraints, line current constraints, voltage fluctuation constraints, harmonic size constraints, and power flow balance constraints; The screening of the gene individuals based on the distribution network operation parameters and pre-established multi-objective functions and constraints includes: Based on each population, determine whether the distribution network node voltage and line current corresponding to each gene individual in the current population meet the voltage deviation constraint and the line current carrying capacity constraint; determine whether the distributed photovoltaic capacity in the current population meets the harmonic size constraint and the voltage fluctuation constraint; if all are satisfied, the fitness of the gene individual is non-zero; otherwise, the fitness of the gene individual is zero; The gene individuals are screened based on the fitness. If the fitness of all gene individuals in the current population is non-zero, the gene individuals in the current population are introduced into the adjacent population according to the set generation number, and the gene individuals with non-zero results in various populations form a mating pool; otherwise, they are not introduced into the adjacent population.

2. The method according to claim 1, wherein The photovoltaic power plants to be connected to the power distribution network are used to construct a set number of populations, and the power distribution network operation parameters corresponding to the individual genes in each population are obtained through Newton-Raphson power flow calculation, including: Based on the photovoltaics to be connected to the distribution network, a set number of photovoltaics are used to construct a genetic individual, and the set number of genetic individuals are used to construct a population; The access capacity of each gene individual is expressed through the decimal code of the photovoltaic genome; The gene individuals are normalized and sequentially brought into Newton-Raphson power flow calculation according to the set access positions to obtain the distribution network operation parameters corresponding to each gene individual.

3. The method according to claim 1, wherein The fitness is calculated by the following formula: Among them, F is the calibration fitness of the k-generation gene individual; f is the calculated value of the objective function of the k-generation gene individual; is the minimum objective function value among the k-generation genetic individuals; ξ k is the k-generation pressure adjustment value, which decreases as k increases.

4. The method according to claim 1, wherein The iterative calculation of the gene individuals obtained by screening to obtain the optimal offspring of each population includes: Step 3-1: Based on the fitness of the genetic individuals, randomly select the father and mother of the genetic individuals in the mating pool through a roulette wheel method, and obtain a random number; Step 3-2: If the random number is less than the set crossover probability, the parent and parent generations are crossovered to obtain the offspring genome; otherwise, no crossover is performed and the parent genome is used as the offspring genome; Step 3-3: If the random number is less than the set mutation probability, a mutation operation is performed on the offspring genome; otherwise, no mutation operation is performed; Step 3-4: Re-execute step 3-1 until each population generates a progeny genome, and select the optimal progeny from each population.

5. An optimization system for photovoltaic access to distribution network, characterized in that: The system comprises: The distribution network operation parameter acquisition module is used to construct a set number of populations based on multiple photovoltaics to be connected to the distribution network, and obtain the distribution network operation parameters corresponding to each gene individual in each population through Newton-Raphson power flow calculation; A screening module, configured to screen the gene individuals based on the distribution network operation parameters and pre-established multi-objective functions and constraints; The optimal offspring screening module is used to iteratively calculate the optimal offspring of each population based on the gene individuals screened; An access scheme confirmation module is used to introduce the optimal offspring into the elite population. If the set iteration end criterion is met, the photovoltaic access scheme represented by the optimal individual gene is obtained; otherwise, the distribution network operation parameter acquisition module is executed again; The distribution network operation parameters obtained in the distribution network operation parameter acquisition module include: distribution network node voltage, line current and distributed photovoltaic capacity; The screening module includes: Objective function construction submodule: The objective function is constructed by maximizing the sum of grid-connected photovoltaic capacity, minimizing operation and maintenance costs, and minimizing grid-connected active power loss; Constraint construction submodule: Constructs constraints based on voltage deviation constraints, line current constraints, voltage fluctuation constraints, harmonic size constraints, and power flow balance constraints; The screening module further includes: Fitness calculation submodule: Based on each population, it determines whether the distribution network node voltage and line current corresponding to each gene individual in the current population meet the voltage deviation constraint and line current carrying capacity constraint; it determines whether the distributed photovoltaic capacity in the current population meets the harmonic size constraint and voltage fluctuation constraint; if all are satisfied, the fitness of the gene individual is non-zero; otherwise, the fitness of the gene individual is zero; Screening submodule: Screen the gene individuals based on the fitness. If the fitness of all gene individuals in the current population is non-zero, the gene individuals in the current population are introduced into the adjacent population according to the set generation number, and the gene individuals with non-zero results in various populations form a mating pool; otherwise, they are not introduced into the adjacent population.

6. The system according to claim 5, wherein: The distribution network operation parameter acquisition module includes: Gene individual construction submodule: based on the photovoltaic power plants to be connected to the distribution network, a gene individual is constructed with a set number of photovoltaic power plants, and a population is constructed with the set number of gene individuals; Access capacity construction submodule: The access capacity of each gene individual is expressed through the decimal code of the photovoltaic genome; The power flow calculation submodule: normalizes the gene individuals and sequentially brings them into the Newton-Raphson power flow calculation according to the set access positions to obtain the distribution network operation parameters corresponding to each gene individual.

7. The system according to claim 5, wherein: The fitness calculation submodule calculates the fitness using the following formula: Among them, F is the calibration fitness of the k-generation gene individual; f is the calculated value of the objective function of the k-generation gene individual; is the minimum objective function value among the k-generation genetic individuals; ξ k is the k-generation pressure adjustment value, which decreases as k increases.

8. The system according to claim 5, wherein: The optimal offspring screening module includes: Father and mother screening submodule: based on the fitness of the gene individuals, randomly select the father and mother from the gene individuals in the mating pool through the roulette method and obtain random numbers; Crossover operation submodule: If the random number is less than the set crossover probability, the parent and parent generations are crossovered to obtain the offspring genome; otherwise, no crossover operation is performed and the parent genome is used as the offspring genome; Mutation operator module: if the random number is less than the set mutation probability, the offspring genome is subjected to mutation operation; otherwise, no mutation operation is performed; Iteration submodule: Rerun the parent and parent screening submodule until each population generates offspring genomes and selects the best offspring from each population.

Images