Distributed photovoltaic and energy storage system dynamic optimization configuration method considering time sequence characteristics

Abstract

The invention relates to a distributed photovoltaic and energy storage system dynamic optimization configuration method considering time sequence characteristics, and the method comprises the steps: constructing a distributed photovoltaic and energy storage configuration optimization model under the time sequence characteristics, and enabling the optimization model to take the minimum total cost of the system as an optimization target; and solving the optimization model by adopting a genetic algorithm, and outputting to obtain an optimal photovoltaic and energy storage configuration scheme. Compared with the prior art, the time sequence characteristic and seasonal change of photovoltaic power generation are comprehensively considered, and the node voltage constraint, the power flow constraint and the distributed photovoltaic and energy storage installation capacity constraint are considered in the constraint conditions of the design optimization model; the optimization target is that the total cost of the system is minimum, the K-means algorithm is utilized to cluster the irradiance data to obtain four typical scenes so as to reduce the calculation cost, and the genetic algorithm is utilized to solve the optimization model, so that the limitation of the PSO algorithm in processing a complex system can be overcome, the problem that the PSO algorithm is easy to fall into a local optimal solution in a complex optimization problem is solved, and the optimization efficiency is improved. And the practicability and effectiveness of an optimization result are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network planning, and in particular to a method for dynamically optimizing the configuration of a distributed photovoltaic and energy storage system taking into account timing characteristics. Background Art

[0002] The randomness and intermittency of photovoltaic power generation lead to low resource utilization. Energy storage has the characteristics of flexible control and rapid response. It is one of the effective means to solve the problem of photovoltaic grid connection and improve consumption. Distributed photovoltaic energy storage configuration scheme is the key to improving the stability of photovoltaic system and energy utilization efficiency.

[0003] In the field of distribution network planning, most existing studies focus on the site selection and capacity determination of distributed power sources in a single scenario, failing to fully consider the timing changes and seasonal fluctuations of distributed photovoltaic power generation, ignoring the complexity of grid operation in actual applications, resulting in the difficulty of implementing the resulting solutions in actual applications. In addition, when discussing distributed photovoltaic integration, existing studies often fail to simultaneously optimize the configuration of energy storage facilities, which may cause energy loss within the system.

[0004] In addition, in terms of algorithm application, the Particle Swarm Optimization (PSO) algorithm is widely used due to its fast convergence characteristics. However, with the expansion of the scale of power systems and the increase in model complexity, the PSO algorithm is prone to fall into local optimality, and its performance is affected by the particle topology structure and is sensitive to the initial settings of the algorithm, resulting in the inability to quickly and accurately solve the optimization results. Summary of the invention

[0005] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a method for dynamic optimization configuration of distributed photovoltaic and energy storage systems taking into account the timing characteristics. The timing characteristics and seasonal changes of photovoltaic power generation are comprehensively considered to construct a distributed photovoltaic and energy storage configuration optimization model, and then a genetic algorithm is used for solution calculation, so as to obtain the optimal configuration optimization result at low cost, quickly and accurately.

[0006] The purpose of the present invention can be achieved by the following technical solution: A method for dynamically optimizing the configuration of a distributed photovoltaic and energy storage system considering timing characteristics, comprising the following steps:

[0007] S1. Construct a distributed photovoltaic and energy storage configuration optimization model under time series characteristics, which takes the minimum total system cost as the optimization goal;

[0008] S2. Use genetic algorithm to solve the optimization model and output the optimal photovoltaic and energy storage configuration plan.

[0009] Furthermore, the optimization model in step S1 includes an objective function and constraints, wherein the objective function includes the investment and operation and maintenance costs of distributed photovoltaic and energy storage, and the operation costs of the power system;

[0010] The constraints include node voltage constraints, power flow constraints, photovoltaic installation capacity constraints and energy storage installation capacity constraints.

[0011] Furthermore, the power system operating costs include online costs, main grid electricity purchase costs and abandoned solar power costs.

[0012] Furthermore, the objective function is specifically:

[0013]

[0014] Among them, TotalCost is the total cost of the system, n PV and n ESS Refers to the number of PV (photovoltaic) and ESS (Energy Storage System), E PV 、E ESS 、E PVloss 、E ext and E NetLoss They are PV power generation, ESS power generation, abandoned photovoltaic power, external power purchase and grid loss power. p ele and p NetLoss They refer to the unit price of PV installation, the unit price of ESS installation, the operation and maintenance cost of PV unit power generation, the operation and maintenance cost of ESS unit power generation, the unit cost of abandoned light, the unit cost of main grid purchased electricity and the unit network loss cost.

[0015] Furthermore, the photovoltaic (PV) installation capacity constraint is specifically:

[0016]

[0017] in, is the number of PV installations at node i, The maximum number of PV installations, L PV Install the candidate node for PV.

[0018] Furthermore, the energy storage system (ESS) installation capacity constraint is specifically:

[0019]

[0020] in, is the number of ESS installed at node i, The maximum number of ESS installations, L ESS Install the candidate nodes for ESS.

[0021] Furthermore, the node voltage constraint is specifically:

[0022]

[0023] in, is the minimum voltage allowed at node i, is the maximum voltage allowed at node i, U i is the voltage at node i.

[0024] Furthermore, the power flow constraint is specifically:

[0025]

[0026] Among them, P is and Q is are the active and reactive injection of nodes i and j respectively, U i is the voltage amplitude at node i, a ij is the element in the i-th row and j-th column of the adjacency matrix composed of the nodes, G ij and B ij is the real and imaginary part of the node admittance matrix, that is, the resistance and reactance of the branch, θ ij is the phase angle difference between nodes i and j.

[0027] Furthermore, step S2 includes the following process:

[0028] S21. Collect radiation data throughout the year;

[0029] S22. Use K-means algorithm to cluster the radiation illumination data and obtain 4 typical scenes, corresponding to the radiation illumination in the four seasons of spring, summer, autumn and winter;

[0030] S23, set the population size, mutation rate, number of elite strategies to be retained, maximum number of iterations, and the maximum number of consecutive times the optimal solution remains unchanged;

[0031] S24, generating an initial group, each individual in the group corresponds to a photovoltaic and energy storage configuration scheme, and is stored in the form of an array, where the elements in the array correspond to the number of PV installations and the number of ESS installations of the node;

[0032] S25, check loop conditions: the number of iterations is less than or equal to the maximum number of iterations and the number of consecutive times the optimal solution remains unchanged is less than the maximum number of times. If so, execute step S26, otherwise execute step S29;

[0033] S26. Check the constraints of each individual in the group at each moment in four typical scenarios. If the constraints are not met, the individual is eliminated.

[0034] S27, calculate the fitness values ​​of all surviving individuals, and retain the individuals with the highest fitness values ​​according to the elite rule;

[0035] S28, perform crossover and mutation operations to generate a new population, and then repeat steps S25 to S27;

[0036] S29. Output the individual with the highest fitness value in the population.

[0037] Furthermore, the common feature of the four typical scenes in step S22 is that the peak values ​​appear at 12 o'clock, 0-5 o'clock and 20-23 o'clock, and the radiation illuminance is close to 0. The difference is that there are significant differences in the values ​​of the peak values, and there is a difference between the earliest time when the radiation illuminance of the four scenes is significantly not 0 and the latest time when it disappears.

[0038] Compared with the prior art, the present invention has the following advantages:

[0039] The present invention first constructs a distributed photovoltaic and energy storage configuration optimization model, which takes the minimum total system cost as the optimization goal, wherein the total system cost is mainly composed of two parts, the first part is the investment and operation and maintenance cost of distributed photovoltaic and energy storage, and the second part is the cost generated by the operation of the power system, which is composed of network loss cost, main network power purchase cost and abandoned light cost; in addition, the setting constraints include node voltage constraint, flow constraint, distributed photovoltaic and energy storage installation capacity constraint. After that, considering the timing characteristics and seasonal changes of photovoltaic power generation, the K-means algorithm is used to cluster the radiation illumination data to obtain 4 typical scenarios, and then the genetic algorithm is used to solve the optimization model, which can effectively reduce the computing cost, overcome the limitations of the PSO algorithm in dealing with complex optimization problems, avoid falling into local optimal solutions, and improve the practicality and effectiveness of the optimization results. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0041] Figure 2 This is a schematic diagram of the process of solving the optimization model in Example 2. DETAILED DESCRIPTION

[0042] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0043] Embodiment 1

[0044] like Figure 1As shown, a method for dynamic optimization configuration of a distributed photovoltaic and energy storage system considering timing characteristics includes the following steps:

[0045] S1. Construct a distributed photovoltaic and energy storage configuration optimization model under time series characteristics, which takes the minimum total system cost as the optimization goal;

[0046] S2. Use genetic algorithm to solve the optimization model and output the optimal photovoltaic and energy storage configuration plan.

[0047] This embodiment applies the above solution to construct a distributed photovoltaic and energy storage configuration optimization model under time sequence characteristics, the main contents of which are:

[0048] The constraints of the optimization model take into account node voltage constraints, power flow constraints, distributed photovoltaic and energy storage installation capacity constraints, etc.

[0049] The optimization goal is to minimize the total cost of the system. The cost is mainly composed of two parts. The first part is the investment and operation and maintenance cost of distributed photovoltaic and energy storage. The second part is the cost generated by the operation of the power system, which is composed of network loss cost, main grid power purchase cost and abandoned light cost.

[0050] Specifically, the objective function of the optimization model is:

[0051]

[0052] Among them, n PV and n ESS Refers to the number of PV and ESS respectively, E PV , E ESS , E PVloss , E ext and E NetLoss They are PV power generation, ESS power generation, abandoned photovoltaic power, external power purchase and grid loss power. p ele and p NetLoss They refer to the unit price of PV installation, the unit price of ESS installation, the operation and maintenance cost of PV unit power generation, the operation and maintenance cost of ESS unit power generation, the unit cost of abandoned light, the unit cost of main grid purchased electricity and the unit network loss cost.

[0053] The constraints are:

[0054] 1) PV (photovoltaic) installation quantity limit

[0055]

[0056] in is the number of PV installations at node i, The maximum number of PV installations, LPV Install the candidate node for PV.

[0057] 2) ESS (Energy Storage System) installation quantity limit

[0058]

[0059] in is the number of ESS installed at node i, The maximum number of ESS installations, L ESS Install the candidate nodes for ESS.

[0060] 3) Node voltage constraints

[0061]

[0062] in is the minimum voltage allowed at node i, is the maximum voltage allowed at node i, U i is the voltage at node i.

[0063] 4) Power flow constraints

[0064]

[0065] Among them, P is and Q is are the active and reactive injection of nodes i and j respectively, U i is the voltage amplitude at node i, a ij is the element in the i-th row and j-th column of the adjacency matrix composed of the nodes, G ij and B ij is the real and imaginary part of the node admittance matrix, that is, the resistance and reactance of the branch, θ ij is the phase angle difference between nodes i and j.

[0066] Embodiment 2

[0067] This embodiment designs a program product, which is used to implement the method steps described in the first embodiment when executed by a computer, such as Figure 2 As shown, the program product uses genetic algorithm to solve the optimization model. The specific process includes:

[0068] 01) Input the annual radiation data in hours;

[0069] 02) The K-means algorithm was used to cluster the radiation illumination data, and four typical scenes were obtained, corresponding to the radiation illumination of the four seasons of spring, summer, autumn and winter. The common characteristics were that the peak value appeared at 12 o'clock, and the radiation illumination was close to 0 from 0 to 5 o'clock and from 20 to 23 o'clock. The difference was that there were significant differences in the peak values, and there was a difference between the earliest time when the radiation illumination of the four scenes was significantly not 0 and the latest time when it disappeared;

[0070] 03) Input network structure, node type, node load and line data to construct a distributed photovoltaic and energy storage configuration optimization model;

[0071] 04) Input population size, mutation rate, number of elite strategies to be retained, maximum number of iterations, and the termination condition for the maximum number of consecutive times the optimal solution remains unchanged;

[0072] 05) Generate an initial group, each individual in the group is a planning scheme, which is stored in the form of an array. The elements in the array correspond to the PV installation quantity and ESS installation quantity of the node. In this embodiment, the planning scheme is an array with a length of 26. The first 16 parameters represent the PV installation quantity of 16 PV candidate nodes, and the last 10 parameters represent the ESS installation quantity of 10 ESS candidate nodes;

[0073] 06) Check the loop condition: the number of iterations is less than or equal to the maximum number of iterations and the number of consecutive times the optimal solution remains unchanged is less than the maximum number of times. If so, execute 07), if not, execute 12);

[0074] 07) Check the constraints of each individual in the group at each moment in four typical scenarios. If the constraints are not met, the individual will be eliminated;

[0075] 08) Calculate the fitness values ​​of all surviving individuals;

[0076] 09) According to the elite rule, retain some individuals with the highest fitness value;

[0077] 10) Perform crossover and mutation operations to generate new populations;

[0078] 11) Repeat steps 06) to 10);

[0079] 12) Output the individual with the highest fitness value in the group.

[0080] In summary, in view of the increasing popularity of distributed photovoltaics in practical applications, this scheme designs and proposes a new distributed photovoltaic and energy storage configuration optimization model, which comprehensively considers the timing characteristics and seasonal changes of photovoltaic power generation, and uses the K-means algorithm to obtain 4 typical scenarios to reduce the computational cost. In order to solve the model, the genetic algorithm (GA) is used to solve it, in order to overcome the limitations of the PSO algorithm in dealing with complex systems, solve the problem that the PSO algorithm is prone to fall into local optimal solutions in complex optimization problems, and improve the practicality and effectiveness of the optimization results.

Claims

1. A method for dynamic optimization configuration of distributed photovoltaic and energy storage systems considering timing characteristics, characterized in that: The following steps are involved: S1. Construct a distributed photovoltaic and energy storage configuration optimization model under time series characteristics, which takes the minimum total system cost as the optimization goal; S2. Use genetic algorithm to solve the optimization model and output the optimal photovoltaic and energy storage configuration plan.

2. According to claim 1, a method for dynamic optimization configuration of distributed photovoltaic and energy storage systems considering timing characteristics is characterized in that: The optimization model in step S1 includes an objective function and constraints, wherein the objective function includes the investment and operation and maintenance costs of distributed photovoltaic and energy storage, and the operation costs of the power system; The constraints include node voltage constraints, power flow constraints, photovoltaic installation capacity constraints and energy storage installation capacity constraints.

3. A method for dynamic optimization configuration of distributed photovoltaic and energy storage systems considering timing characteristics according to claim 2, characterized in that: The power system operating costs include online costs, main grid power purchase costs and abandoned solar power costs.

4. A method for dynamic optimization configuration of distributed photovoltaic and energy storage systems considering timing characteristics according to claim 3, characterized in that: The objective function is specifically: Among them, TotalCost is the total cost of the system, n PV and n ESS Refers to the number of PV photovoltaic and ESS energy storage systems, E PV 、E ESS 、E PVliss 、E ext and E NetLoss They are PV power generation, ESS power generation, abandoned photovoltaic power, external power purchase and grid loss power. p ele and p NetLoss They refer to the unit price of PV installation, the unit price of ESS installation, the operation and maintenance cost of PV unit power generation, the operation and maintenance cost of ESS unit power generation, the unit cost of abandoned light, the unit cost of main grid purchased electricity and the unit network loss cost.

5. A method for dynamic optimization configuration of distributed photovoltaic and energy storage systems considering timing characteristics according to claim 2, characterized in that: The photovoltaic PV installation capacity constraint is specifically: in, is the number of PV installations at node i, The maximum number of PV installations, L PV Install the candidate node for PV.

6. A method for dynamic optimization configuration of distributed photovoltaic and energy storage systems considering timing characteristics according to claim 2, characterized in that: The energy storage ESS installation capacity constraint is specifically: in, is the number of ESS installed at node i, The maximum number of ESS installations, L ESS Install the candidate nodes for ESS.

7. A method for dynamic optimization configuration of distributed photovoltaic and energy storage systems considering timing characteristics according to claim 2, characterized in that: The node voltage constraint is specifically: in, is the minimum voltage allowed at node i, is the maximum voltage allowed at node i, U i is the voltage at node i.

8. A method for dynamic optimization configuration of distributed photovoltaic and energy storage systems considering timing characteristics according to claim 2, characterized in that: The power flow constraints are specifically: Among them, P is and Q is are the active and reactive injection of nodes i and j respectively, U i is the voltage amplitude at node i, a ij is the element in the i-th row and j-th column of the adjacency matrix composed of the nodes, G ij and B ij is the real and imaginary part of the node admittance matrix, that is, the resistance and reactance of the branch, θ ij is the phase angle difference between nodes i and j.

9. A method for dynamic optimization configuration of distributed photovoltaic and energy storage systems considering timing characteristics according to claim 1, characterized in that: The step S2 includes the following process: S21. Collect radiation data throughout the year; S22. Use K-means algorithm to cluster the radiation illumination data and obtain 4 typical scenes, corresponding to the radiation illumination in the four seasons of spring, summer, autumn and winter; S23, set the population size, mutation rate, number of elite strategies to be retained, maximum number of iterations, and the maximum number of consecutive times the optimal solution remains unchanged; S24, generating an initial group, each individual in the group corresponds to a photovoltaic and energy storage configuration scheme, and is stored in the form of an array, where the elements in the array correspond to the number of PV photovoltaic installations and the number of ESS energy storage system installations of the node; S25, check loop conditions: the number of iterations is less than or equal to the maximum number of iterations and the number of consecutive times the optimal solution remains unchanged is less than the maximum number of times. If so, execute step S26, otherwise execute step S29; S26. Check the constraints of each individual in the group at each moment in four typical scenarios. If the constraints are not met, the individual is eliminated. S27, calculate the fitness values ​​of all surviving individuals, and retain the individuals with the highest fitness values ​​according to the elite rule; S28, perform crossover and mutation operations to generate a new population, and then repeat steps S25 to S27; S29. Output the individual with the highest fitness value in the population.

10. A method for dynamic optimization configuration of distributed photovoltaic and energy storage systems considering timing characteristics according to claim 9, characterized in that: The common feature of the four typical scenes in step S22 is that the peak values ​​appear at 12 o'clock, 0-5 o'clock and 20-23 o'clock, and the radiation illuminance is close to 0. The difference is that there are significant differences in the values ​​of the peak values, and there is a difference between the earliest time when the radiation illuminance of the four scenes is significantly not 0 and the latest time when it disappears.

Images