Distributed photovoltaic microgrid system optimization method considering demand side response and energy storage

The support vector machine model and energy storage model improved by particle swarm algorithm, combined with the random weighting method, optimize the photovoltaic microgrid system, solve the uncertainty and intermittent problems of photovoltaic power generation, and realize the maximization of photovoltaic absorption capacity and operating cost Minimizing the light abandonment rate reduces the feasibility of photovoltaic applications.

CN119029948AActive Publication Date: 2024-11-26NORTH CHINA ELECTRIC POWER UNIV +1

Patent Information

Application Number
CN202410893309.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-04
Publication Date
2024-11-26
Estimated Expiration
2044-07-04

AI Technical Summary

Technical Problem

The uncertainty and intermittent nature of photovoltaic power generation lead to fluctuations in the grid voltage and frequency, resulting in light abandonment. The existing technology is difficult to effectively promote the grid connection and absorption of photovoltaic power generation, affecting the safe and stable operation of the distribution network.

Method used

Establish a photovoltaic power prediction model for supporting vector machines based on improved particle swarm algorithm, combine energy storage and load models, optimize the photovoltaic microgrid system through a random weighting method, maximize the photovoltaic absorption capacity and minimize operating costs, and consider the distribution network Economicality and photovoltaic consumption capacity.

Benefits of technology

It effectively reduces the abandonment rate of photovoltaic power plants, increases the photovoltaic absorption rate, provides feasibility guidance for photovoltaic applications, and improves the feasibility and economicality of photovoltaic applications in the region.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119029948A_ABST
    Figure CN119029948A_ABST
Patent Text Reader

Abstract

The invention belongs to the technical field of photovoltaic micro-grid control, and particularly relates to a distributed photovoltaic micro-grid system optimization method considering demand side response and energy storage, and the method comprises the steps: A, building a photovoltaic day-ahead output prediction model; b, establishing an energy storage and load model; and step C, establishing a photovoltaic consumption capability evaluation optimization model. And predicting the photovoltaic day-ahead output by using a PSO-SVM regression prediction method. Aiming at the problem that the photovoltaic power generation light abandoning phenomenon is serious, the influence of energy storage and demand response on the photovoltaic consumption rate is considered, and an optimization target is established by considering the maximum photovoltaic consumption capacity and the minimum network active power loss from the constraint conditions that the node voltage is not out of limit, the power flow is not overloaded, the energy storage is not out of limit and the like. And a distributed photovoltaic consumption capability evaluation optimization model is formed, and the consumption capability and economy of the photovoltaic system are comprehensively considered. Simulation results show that the model can effectively reduce the light abandoning rate of the photovoltaic power station, and provides feasibility guidance for photovoltaic application in a region.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of photovoltaic microgrid control, and in particular relates to a distributed photovoltaic microgrid system optimization method considering demand-side response and energy storage. Background Art

[0002] With the increasing shortage of energy and the deteriorating environment, the development and application of renewable energy have become the focus of many research topics. Photovoltaic power generation, as the most abundant form of power generation among new energy sources and with fewer geographical restrictions, has developed particularly rapidly. However, due to the uncertainty and intermittency of photovoltaics, their voltage and frequency are affected after they are connected to the grid, making it difficult for the power quality of the grid to meet the requirements. It can be seen that the large-scale access of photovoltaics has brought many challenges to the safe and stable operation of the distribution network. When the power grid cannot absorb large-scale photovoltaic grid-connected power, it has to face a relatively serious phenomenon of abandoned light. Therefore, at this stage, research on promoting the grid connection and absorption of photovoltaic power generation has extremely important application value.

[0003] Energy storage units can store electricity when there is excess power and supply it when there is insufficient power. Microgrids can use demand response to regulate user electricity consumption through electricity market prices. Energy storage and demand response are two means of regulation that can change the photovoltaic absorption capacity in microgrids. Researching and analyzing the specific impacts of the two is also of great practical significance.

[0004] Therefore, there is an urgent need for an optimization method for distributed photovoltaic microgrid systems that takes demand-side response and energy storage into consideration, an optimization scheme for distributed photovoltaic microgrid systems that takes demand-side response and energy storage into consideration, and a photovoltaic microgrid system absorption optimization scheme that fully considers the energy storage characteristics and load characteristics, and comprehensively considers the photovoltaic absorption capacity of the distribution network and the economy of the distribution network operation. Summary of the invention

[0005] The object of the present invention is to provide a distributed photovoltaic microgrid system optimization method considering demand-side response and energy storage, characterized in that it comprises the following steps:

[0006] Step A: Establish a photovoltaic day-ahead output prediction model, use the photovoltaic power prediction method based on particle swarm algorithm to improve the support vector machine, and process the statistical time data of photovoltaic output through direct prediction method on the basis of physical analysis of meteorological influencing factors of photovoltaic power generation. Then, modify the prediction results under different external weather conditions according to the statistical rules of historical data.

[0007] Step B: Establishing an energy storage and load model, the energy storage and load model includes: an energy storage model and a load absorption energy consumption model; the energy storage model realizes peak shaving and valley filling while suppressing the anti-load characteristics of new energy power generation; the load absorption energy consumption model adjusts the operation time of the equipment according to the photovoltaic output characteristics, transfers the time-shiftable load during the peak power consumption period to the low power consumption period or the peak period of photovoltaic output, and performs energy consumption analysis according to the proportion of different load categories;

[0008] Step C: Establish an optimization model for the photovoltaic microgrid system with energy storage system and demand response, measure the photovoltaic absorption capacity of the distribution network by the sum of the distributed photovoltaic access capacity of each node, and use the random weighting method to weight the two objective functions of maximum photovoltaic absorption capacity and minimum active network loss, and integrate them into the optimization goal. Combined with the power flow constraints, node voltage deviation constraints, distributed photovoltaic power generation energy constraints and energy storage system energy upper limit constraints, the photovoltaic absorption capacity and economy of the distribution network are considered to achieve the optimization of the distributed photovoltaic microgrid system.

[0009] The step A of establishing a photovoltaic day-ahead output prediction model includes the following steps:

[0010] Step A1: Based on the least squares support vector machine, the inequality constraints in the optimization formula are converted into equality constraints; based on the data sample set, the least squares support vector machine regression algorithm is used, and the nonlinear mapping φ(x) is used to project the input data into a high-dimensional feature space, so that the function value y corresponding to the sample x is approximately expressed as f(x);

[0011] y=f(x)=ωφ(x)+b (1)

[0012] In the formula, ω is the weight vector in the feature space, and b is the bias;

[0013] The data sample set is: D = {(x i ,y i )|(i=1,2,...,l)}, where: D represents the data sample set, l is the number of samples in the sample set, x i is the i-th sample data, y i is the function value corresponding to the i-th sample data;

[0014] The loss function of the photovoltaic day-ahead output prediction model is defined as:

[0015]

[0016] In the formula, ε is the insensitive coefficient, e is the loss function, f(x i ) is the function value corresponding to the i-th sample;

[0017] The least squares support vector regression is expressed as an optimization problem for equation (2):

[0018]

[0019] In the formula, γ is the penalty factor, J(ω,e) is the least squares support vector, φ(x i ) is x i The nonlinear mapping, e i is the sample error calculated by formula (2), y i Represents the function value corresponding to the i-th sample data;

[0020] The Lagrangian function is established based on formula (3), namely:

[0021]

[0022] In the formula, a i is the Lagrange multiplier, L is the Lagrange function, a represents the set of Lagrange multipliers, e i is the sample error calculated by formula (2);

[0023] According to the KKT condition, the linear equation is:

[0024]

[0025] Where I is the l×l unit matrix, γ -1 is the inverse of the penalty factor, b is the bias, and the remaining expressions are:

[0026]

[0027] In the formula, e l is the identity matrix, Ω i,j For x i With x j The product of the nonlinear mapping, a is the Lagrange multiplier matrix, y is the sample function value matrix, K(x i ,x j ) is the kernel function, T represents transpose;

[0028] Solve the linear equations in equation (6) to obtain the values ​​of a and b, and obtain the regression fitting function, which is:

[0029]

[0030] In the formula, K(x,x i ) is a kernel function, and the kernel function adopts a radial basis kernel function, and the radial basis kernel function is:

[0031]

[0032] Where σ is the width of the radial basis kernel function;

[0033] Step A2: performing optimization based on a particle swarm optimization algorithm and outputting an optimization result, wherein the optimization result includes: an optimal penalty parameter c and an optimal kernel function parameter g;

[0034] Suppose in a D-dimensional search space, there are n particles forming a population M = (M1, M2, ..., M n ), where the position of the i-th particle is X i =(X i1 ,X i2 ,...,X iD ) T , speed is V i =(V i1 ,V i2 ,...,V iD ) T , the individual extreme value is P i =(P i1 ,P i2 ,...,P iD ) T , the global extreme value of the population is P g =(P g1 ,P g2 ,...,P gD ) T ;

[0035] The evolution mode of the particle swarm optimization algorithm is:

[0036] Particle swarm algorithm evolution model = particle historical inertia + particle self-cognition + particle group cognition

[0037] The update formula of velocity and position in the particle iterative optimization process is expressed as:

[0038] V id (k+1)=ωV id (k)+c1r1[P id (k)-X id (k)]+c2r2[P gd (k)-X id (k)] (9)

[0039] X id (k+1)=X id (k)+V id (k+1) (10)

[0040] Where V id (k) is the optimization speed of the current particle, P id (k) is the optimal position of the current particle, X id (k) Current particle position, Pgd (k) is the optimal position of the current particle group, ω is the inertia factor, which is mainly used to balance the global search and local search capabilities; d = 1, 2, ..., n, d represents the number of iterations; i = 1, 2, ..., D, i represents the i-th particle; k is the evolutionary generation; c1 and c2 are acceleration factors, which are non-negative values, c1 represents the degree of inheritance of the particle to its own optimal position, and c2 represents the degree of inheritance of the particle to the global optimal position; r1 and r2 are random numbers, and the value space is [0, 1];

[0041] Each particle determines the adaptive value based on the objective function during each iteration and determines the optimal position P of the current particle. id (k) and the optimal position of the group P gd (k), update the speed and position of each particle through equation (9) and equation (10); when the number of iterations reaches the set value or the result meets the accuracy requirement, the optimization is terminated, otherwise the iteration continues.

[0042] The photovoltaic power generation prediction method based on the particle swarm algorithm to improve the support vector machine in step A includes the following steps:

[0043] Import historical photovoltaic data;

[0044] Use particle swarm algorithm for optimization;

[0045] Determine whether the optimization conditions are met;

[0046] If the optimization condition is not met, update the historical optimal position of each particle; update the global optimal position of the group;

[0047] If the optimization condition is met, the optimization result is output, and the optimization result includes: the best penalty parameter c and the best kernel function parameter g;

[0048] Import the optimization results into support vector regression for training and prediction;

[0049] Calculate mean square error and coefficient of determination;

[0050] Output predicted power.

[0051] The step B of establishing the energy storage and load model comprises the following steps:

[0052] Step B1: Said The energy storage model is based on the power supply characteristics and load characteristics of the energy storage system. It can store new energy electricity in the form of load during the off-peak period; during the peak period, it releases the electricity stored during the off-peak period to relieve the power supply pressure of the power system; charging and discharging cannot be carried out at the same time in the same period, and the charging power and discharge power The following constraints should be met at time t:

[0053]

[0054] When the distributed photovoltaic output is greater than the load, When the distributed photovoltaic output is less than the load,

[0055] Step B2: The load absorption energy consumption model calculates the load absorption of photovoltaic power in the domain. By coordinating the time-shiftable load and the fixed load, the energy consumption analysis is performed according to the proportion of different load categories. The energy consumption calculation method of load absorption is as follows:

[0056] C ec =p c ∫t c L c (t)dt+E c (12)

[0057] In the formula, C ec is the energy consumption of load absorption, p c It is the time-of-use electricity price, which includes the electricity prices in peak, flat and valley periods; t c is the load operation time; L c is the load power; E c For running costs.

[0058] , the calculation method of photovoltaic absorption capacity in step C is:

[0059]

[0060] Where f1 is the photovoltaic absorption capacity; N PV P is the total number of nodes connected to the photovoltaic system; PV,m is the photovoltaic access capacity of the mth node.

[0061] The calculation method of active network loss in step C is:

[0062]

[0063] Where f2 is the active network loss, {N B} is the set of all nodes in the system; P ij ,Q ij are respectively the active power and reactive power flowing through branch ij; U i is the voltage value of node i; R ij is the line resistance between nodes i and j.

[0064] The function expression of the random weighting method in step C is:

[0065] maxf=ω1f1-ω2f2 (15)

[0066] Where f is the total objective function, ω1, ω2 are random numbers, and ω1+ω2=1, ω1≤1, ω2≤1.

[0067] The beneficial effects of the present invention are:

[0068] The present invention establishes an optimization scheme for a distributed photovoltaic microgrid system that takes into account demand-side response and energy storage. First, the photovoltaic day-ahead output is predicted using the particle swarm optimization for support vector machine (PSO-SVM) regression prediction method. In view of the serious problem of photovoltaic power generation abandonment, based on the constraints of node voltage not exceeding the limit, current not exceeding the limit, and energy storage not exceeding the limit, a distributed photovoltaic absorption capacity optimization model is established with the goal of maximizing photovoltaic absorption capacity and optimizing economy. The impact of energy storage system and demand-side response on photovoltaic absorption rate is analyzed by simulation. The simulation results show that the model can effectively reduce the abandonment rate of photovoltaic power stations and provide feasibility guidance for photovoltaic applications in the region.

[0069] The present invention uses an intelligent algorithm to predict the day-ahead photovoltaic output, fully considers the energy storage characteristics and load characteristics, comprehensively considers the photovoltaic absorption capacity of the distribution network and the economy of the distribution network operation, takes the dual objectives of maximizing the system photovoltaic absorption capacity and minimizing the operating cost, combines equality and inequality constraints, establishes a photovoltaic absorption capacity optimization model, and proposes a photovoltaic microgrid system absorption optimization plan. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 A flow chart of a distributed photovoltaic microgrid system optimization method considering demand-side response and energy storage according to the present invention;

[0071] Figure 2 This is a flow chart of the distributed photovoltaic output prediction method;

[0072] Figure 3 It is the node structure diagram of the simulation system;

[0073] Figure 4 It is a schematic diagram of the fitness change curve of the particle swarm;

[0074] Figure 5 Comparison chart between the predicted results and the actual results of the test set;

[0075] Figure 6 Schematic diagram of photovoltaic output and typical daily load curve, where (a) is the curve without adding energy storage, and (b) is the curve with adding energy storage;

[0076] Figure 7Schematic diagrams of output curves of four photovoltaic power stations, where (a) is the output curve of photovoltaic power station 2, (b) is the output curve of photovoltaic power station 3, (c) is the output curve of photovoltaic power station 4, and (d) is the output curve of photovoltaic power station 8;

[0077] Figure 8 Schematic diagram of energy change curve of energy storage system installed at 4 photovoltaic nodes;

[0078] Fig. 9 Schematic diagram of photovoltaic output, energy storage system output and upper-level bus power curve. DETAILED DESCRIPTION

[0079] The present invention provides a distributed photovoltaic microgrid system optimization method considering demand-side response and energy storage, and the present invention is further described in detail below in conjunction with the accompanying drawings.

[0080] The present invention proposes an optimization scheme for a distributed photovoltaic microgrid system that takes into account energy storage systems and demand-side response, uses an artificial intelligence algorithm to predict the photovoltaic day-ahead output, and establishes a photovoltaic day-ahead output prediction model, an energy storage and load model, and a photovoltaic microgrid system optimization model that takes into account energy storage systems and demand response.

[0081] like Figure 1 The embodiment of the present invention shown discloses a distributed photovoltaic microgrid system optimization method considering demand-side response and energy storage, comprising the following steps:

[0082] Step A: Establish a photovoltaic day-ahead output prediction model, use the photovoltaic power prediction method based on particle swarm algorithm to improve the support vector machine, and process the statistical time data of photovoltaic output through direct prediction method on the basis of physical analysis of meteorological influencing factors of photovoltaic power generation. Then, modify the prediction results under different external weather conditions according to the statistical rules of historical data.

[0083] In this embodiment, a photovoltaic day-ahead output prediction model is established. A photovoltaic power prediction method based on particle swarm optimization for support vector machine (PSO-SVM) analyzes the photovoltaic power generation system from a physics perspective. On the basis of analyzing the meteorological influencing factors of photovoltaic power generation, the direct prediction method is combined to process the statistical time data of photovoltaic output. The prediction results under different external weather conditions are corrected through the statistical rules of historical data, which reduces the dependence of the predicted photovoltaic power generation on meteorological data and improves the prediction efficiency and accuracy of photovoltaic power generation. The specific process is as follows: Figure 2 shown.

[0084] The photovoltaic power generation prediction method based on the particle swarm algorithm to improve the support vector machine in step A includes the following steps:

[0085] Import historical photovoltaic data;

[0086] Use particle swarm algorithm for optimization;

[0087] Determine whether the optimization conditions are met;

[0088] If the optimization condition is not met, update the historical optimal position of each particle; update the global optimal position of the group;

[0089] If the optimization condition is met, the optimization result is output, and the optimization result includes: the best penalty parameter c and the best kernel function parameter g;

[0090] Import the optimization results into support vector regression for training and prediction;

[0091] Calculate mean square error and coefficient of determination;

[0092] Output predicted power.

[0093] The step A of establishing a photovoltaic day-ahead output prediction model includes the following steps:

[0094] Step A1: Based on the least squares support vector machine, the inequality constraints in the optimization formula are converted into equality constraints; based on the data sample set, the least squares support vector machine regression algorithm is adopted, and the input data is projected into a high-dimensional feature space using the nonlinear mapping φ(x), so that the function value y corresponding to the sample x is approximately represented by f(x).

[0095] In this embodiment, the least squares support vector machine converts the inequality constraints in the optimization formula into equality constraints, thereby greatly reducing the computational complexity and improving the efficiency of solving the regression prediction function. The regression algorithm based on the least squares support vector machine is referred to as the least squares support vector regression (LS-SVR). For the data sample set D = {(x i ,y i )|(i=1,2,...,l)}where: D represents the data sample set, l is the number of samples in the sample set, x i is the i-th sample data, y i is the function value corresponding to the i-th sample data; the support vector machine uses nonlinear mapping φ(x) to project the input data into a high-dimensional feature space, so that the function value y corresponding to the sample x can be approximately expressed by f(x);

[0096] y=f(x)=ωφ(x)+b (1)

[0097] Where ω is the weight vector in the feature space and b is the bias.

[0098] The loss function of the photovoltaic day-ahead output prediction model is defined as:

[0099]

[0100] In the formula, ε is the insensitive coefficient, e is the loss function, f(x i ) is the function value corresponding to the i-th sample.

[0101] Therefore, LS-SVR can be formulated as an optimization problem for the above function.

[0102] The least squares support vector regression is expressed as an optimization problem for equation (2):

[0103]

[0104] In the formula, γ is the penalty factor. Taking an appropriate value can prevent the model from overfitting or underfitting problems. J(ω,e) is the least squares support vector, φ(x i ) is x i The nonlinear mapping, e i is the sample error calculated by formula (2), y i Represents the function value corresponding to the i-th sample data;

[0105] The equality constraint in the optimization formula is shown in formula (3).

[0106] Therefore, the Lagrangian function can be established by equation (3), namely:

[0107]

[0108] In the formula, a i is the Lagrange multiplier, L is the Lagrange function, a represents the set of Lagrange multipliers, e i is the sample error calculated by formula (2).

[0109] According to the KKT (Ksrush-Kuhn-Tucker) condition, the linear equation is:

[0110]

[0111] Where I is the l×l unit matrix, γ -1 is the inverse of the penalty factor, b is the bias, and the remaining expressions are:

[0112]

[0113] In the formula, e lis the identity matrix, Ω i,j For x i With x j The product of the nonlinear mapping, a is the Lagrange multiplier matrix, y is the sample function value matrix, K(x i ,x j ) is the kernel function, T represents transpose;

[0114] Solving the linear equations in equation (6) can obtain the values ​​of a and b, and obtain the regression fitting function;

[0115] The regression fitting function is:

[0116]

[0117] In the formula, K(x,x i ) is the kernel function, and the radial basis function (RBF) kernel function is used, and the expression is:

[0118]

[0119] Where σ is the width of the radial basis kernel function.

[0120] Step A2: performing optimization based on a particle swarm optimization algorithm and outputting an optimization result, wherein the optimization result includes: an optimal penalty parameter c and an optimal kernel function parameter g;

[0121] In this embodiment, the particle swarm optimization algorithm is used to select parameters. The following is its basic algorithm.

[0122] Suppose in a D-dimensional search space, there are n particles forming a population M = (M1, M2, ..., M n ), where the position of the i-th particle is X i =(X i1 ,X i2 ,...,X iD ) T , speed is V i =(V i1 ,V i2 ,...,V iD ) T , the individual extreme value is P i =(P i1 ,P i2 ,...,P iD ) T , the global extreme value of the population is P g =(P g1 ,P g2 ,...,P gD ) T ;

[0123] The speed formula of the particle swarm optimization algorithm consists of three parts, which are usually expressed as follows:

[0124] The evolution mode of the particle swarm optimization algorithm is:

[0125] Particle swarm algorithm evolution model = particle historical inertia + particle self-cognition + particle group cognition

[0126] Therefore, the update formula of the velocity and position of the particle in the iterative optimization process is expressed as:

[0127] V id (k+1)=ωV id (k)+c1r1[P id (k)-X id (k)]+c2r2[P gd (k)-X id (k)] (9)

[0128] X id (k+1)=X id (k)+V id (k+1) (10)

[0129] Where V id (k) is the optimization speed of the current particle, P id (k) is the optimal position of the current particle, X id (k) Current particle position, P gd (k) is the optimal position of the current particle group, ω is the inertia factor, which is mainly used to balance the global search and local search capabilities; d = 1, 2, ..., n, d represents the number of iterations; i = 1, 2, ..., D, i represents the i-th particle; k is the evolutionary generation; c1 and c2 are acceleration factors, which are non-negative values, c1 represents the degree of inheritance of the particle to its own optimal position, and c2 represents the degree of inheritance of the particle to the global optimal position; r1 and r2 are random numbers, and the value space is [0, 1];

[0130] Each particle needs to determine the adaptive value based on the objective function during each iteration to determine the optimal position P of the current particle. id (k) and the optimal position of the group P gd (k), update the velocity and position of each particle through equation (9) and equation (10). When the number of iterations reaches the set value or the result meets the accuracy requirement, the optimization is terminated, otherwise the iteration continues.

[0131] Step B: Establishing an energy storage and load model, the energy storage and load model includes: an energy storage model and a load absorption energy consumption model; the energy storage model realizes peak shaving and valley filling while suppressing the anti-load characteristics of new energy power generation; the load absorption energy consumption model adjusts the operation time of the equipment according to the photovoltaic output characteristics, transfers the time-shiftable load during the peak power consumption period to the low power consumption period or the peak period of photovoltaic output, and performs energy consumption analysis according to the proportion of different load categories;

[0132] In this embodiment, an energy storage and load model is established. The energy storage characteristics are fully considered to establish the energy storage model. The operation time of the equipment is adjusted according to the photovoltaic output characteristics, and the time-shiftable load during the peak power consumption period is transferred to the low power consumption period or the peak period of photovoltaic output. The energy consumption analysis is performed according to the proportion of different load categories, and a calculation model for load absorption and energy consumption is established.

[0133] Step B1: Said The energy storage model is based on the power supply characteristics and load characteristics of the energy storage system. It can store new energy electricity in the form of load during the off-peak period; during the peak period, it releases the electricity stored during the off-peak period to alleviate the power supply pressure of the power system;

[0134] In this embodiment, the energy storage system has flexible power throughput characteristics, which can effectively reduce peak loads and fill valleys while suppressing the reverse load characteristics of renewable energy generation, promote the consumption of renewable energy, and ensure the safety and stability of the main grid. In addition to power supply characteristics, the energy storage system also has load characteristics: it can store renewable energy electricity in the form of loads during periods of low power consumption; during peak power consumption, it can release the electricity stored during valleys to relieve the power supply pressure of the power system.

[0135] Charging and discharging cannot be performed at the same time in the same period of time, that is, the charging power and discharge power At time t, the following conditions should be met:

[0136]

[0137] When the distributed photovoltaic output is greater than the load, When the distributed photovoltaic output is less than the load,

[0138] Step B2: The load absorption energy consumption model calculates the load absorption of photovoltaic power in the domain, and performs energy consumption analysis according to the proportion of different load categories through the coordination of time-shiftable loads and fixed loads;

[0139] In this embodiment, the load absorption of photovoltaic in the domain is mainly through the coordination of time-shiftable load and fixed load, adjusting the operation time of the equipment according to the photovoltaic output characteristics, and transferring the time-shiftable load during the peak power consumption period to the low power consumption period or the peak period of photovoltaic output.

[0140] Energy consumption analysis is performed based on the proportion of different load categories, and the energy consumption of load absorption is calculated as follows:

[0141] C ec =p c ∫t c L c (t)dt+E c (12)

[0142] In the formula, p c It is the time-of-use electricity price, which includes the electricity prices in peak, flat and valley periods; t c is the load operation time; L c is the load power; E c For running costs.

[0143] Step C: Establish an optimization model for the photovoltaic microgrid system with energy storage system and demand response, measure the photovoltaic absorption capacity of the distribution network by the sum of the distributed photovoltaic access capacity of each node, and use the random weighting method to weight the two objective functions of maximum photovoltaic absorption capacity and minimum active network loss, and integrate them into the optimization goal. Combined with the power flow constraints, node voltage deviation constraints, distributed photovoltaic power generation energy constraints and energy storage system energy upper limit constraints, the photovoltaic absorption capacity and economy of the distribution network are considered to achieve the optimization of the distributed photovoltaic microgrid system.

[0144] In this embodiment, an optimization model of a photovoltaic microgrid system with energy storage system and demand response is established. The sum of the distributed photovoltaic access capacity of each node is used to measure the photovoltaic absorption capacity of the distribution network, and the maximum absorption capacity and the minimum active network loss are weighted and integrated as the optimization target. Combined with the power flow constraint, node voltage deviation constraint, distributed photovoltaic power generation energy constraint and energy storage system energy upper limit constraint, the photovoltaic absorption capacity and economy of the distribution network are considered.

[0145] The sum of the distributed photovoltaic access capacity of each node is taken as the photovoltaic absorption capacity of the distribution network. The calculation method of the photovoltaic absorption capacity in step C is:

[0146]

[0147] Where N PV P is the total number of nodes connected to the photovoltaic system; PV,m is the photovoltaic access capacity of the mth node.

[0148] After photovoltaic access, the calculation method of active network loss in step C is:

[0149]

[0150] Where f2 is the active network loss, {N B} is the set of all nodes in the system; Pij ,Q ij are respectively the active power and reactive power flowing through branch ij; U i is the voltage value of node i; R ij is the line resistance between nodes i and j.

[0151] In order to comprehensively consider the photovoltaic absorption capacity of the distribution network and the economic efficiency of the distribution network operation, in this embodiment, the function expression of the random weighted method in step C is:

[0152] maxf=ω1f1-ω2f2 (15)

[0153] Where f is the total objective function, ω1, ω2 are random numbers, and ω1+ω2=1, ω1≤1, ω2≤1.

[0154] In order to enable those skilled in the art to better understand the present invention and appreciate the advantages of the present invention over the prior art, the applicant further explains the present invention in conjunction with specific embodiments.

[0155] 1. Simulation system node structure diagram, particle swarm fitness change curve, photovoltaic prediction result curve

[0156] The node structure of the simulation system is as follows: Figure 3 As shown in the figure, the power generation is predicted using the newly opened PV data of Guoneng Daily. There are 66,860 data sets in total, from April 1, 2016 to April 30, 2018, with a sampling interval of every 15 minutes. Among them, 50,000 data sets are randomly extracted as training sets, and 50 data sets are randomly extracted from the remaining samples as training sets for experimental comparison.

[0157] In the process of using the particle swarm algorithm to optimize the support vector machine parameters, the maximum number of iterations is 100, the number of particle swarms n = 50, the acceleration factors c1 = 1.2, c2 = 1.2, the search range of the penalty parameter c is [0.001, 50], and the search range of the kernel function parameter g is [0.001, 20]. The particle swarm fitness (root mean square error) changes as follows Figure 4 As shown in the figure, the optimal penalty parameter c is 15.8334 and the kernel function parameter g is 0.05591.

[0158] The relative standard deviation (RSD) is used to measure the repeatability of regression prediction, that is, the degree of consistency between independent prediction results. The calculation formula is:

[0159]

[0160] In the formula, f RSD is the relative standard deviation, x i is the predicted value, is the predicted average value, and n is the number of predictions.

[0161] The root mean square error (MSE) is used to evaluate the accuracy of regression prediction, that is, the degree of consistency between the predicted result and the true value. The calculation formula is:

[0162]

[0163] In the formula, f MSE is the root mean square error, is the actual value of photovoltaic output, x i is the predicted value of photovoltaic output. The operation results are as follows Figure 5 As shown in the figure, the prediction results of the support vector machine regression test set show that the correlation coefficient is 0.90375 and the root mean square error is 0.69491, which guarantees a high prediction accuracy for photovoltaic output.

[0164] 2. Analysis of photovoltaic power consumption capacity

[0165] In order to study the impact of energy storage systems on microgrid photovoltaic consumption, two scenarios are set up for comparative optimization analysis, one without adding energy storage systems and the other with adding energy storage systems. Figure 6 As shown, (a) is a schematic diagram of the curve without adding energy storage, and (b) is a schematic diagram of the curve when adding energy storage.

[0166] from Figure 6 It can be clearly seen that when the distribution system does not add energy storage, Figure 6 As shown in (a), there is an obvious mismatch between the supply and demand relationship of photovoltaic power generation and load. The photovoltaic power generation output exceeds the load from 9 to 17 o'clock, the load peaks at around 20 o'clock, and the photovoltaic output is 0 after 20 o'clock. After adding the energy storage system, Figure 6 As shown in (b), the photovoltaic power generation and load are basically matched at 9-17 o'clock, and the photovoltaic power generation is absorbed to a greater extent. The comparison of the photovoltaic power generation absorption rate of the four photovoltaic power stations is shown in Table 1.

[0167] Table 1 Comparison of photovoltaic consumption rate before and after optimization

[0168]

[0169] Comparing the two scenarios, we can see that the introduction of the energy storage system has increased the photovoltaic absorption rate of the microgrid from the original 45.0454% to 78.1356%, which has greatly improved the photovoltaic absorption capacity. Figure 7 As shown, (a) is the output curve of photovoltaic power station 2, (b) is the output curve of photovoltaic power station 3, (c) is the output curve of photovoltaic power station 4, and (d) is the output curve of photovoltaic power station 8.

[0170] The energy changes of the energy storage system installed at the four photovoltaic nodes are as follows: Figure 8 As shown in the figure, from 9:00 to 17:00, since the photovoltaic output can fully supply the system load, the photovoltaic node output is in oversupply. At this time, the energy storage system can store new energy electricity in the form of load; around 20:00, due to the lack of light intensity, the photovoltaic output is zero, the photovoltaic node output cannot meet the system load demand, and the system output is in short supply. At this time, the energy storage system can release the stored electricity to alleviate the power supply pressure of the power system, and the energy storage system exhibits power supply characteristics.

[0171] The photovoltaic output, energy storage system output and upper bus power are as follows: Fig. 9 As shown in the figure, the entire distribution network system balances the active load through the output of photovoltaic power and the output of the energy storage system. Starting from 10:00, the sum of the output of photovoltaic power and the output of the energy storage system in the distribution system can fully meet the active load demand. The system does not need to obtain active power from the upper bus, which means that in theory, it can realize the local control of distributed photovoltaic power consumption. The local control layer works quickly, which can maximize the photovoltaic consumption rate while minimizing the operating cost.

Claims

1. A distributed photovoltaic microgrid system optimization method considering demand-side response and energy storage, characterized in that: The following steps are involved: Step A: Establish a photovoltaic day-ahead output prediction model, use the photovoltaic power prediction method based on particle swarm algorithm to improve the support vector machine, and process the statistical time data of photovoltaic output through direct prediction method on the basis of physical analysis of meteorological influencing factors of photovoltaic power generation. Then, modify the prediction results under different external weather conditions according to the statistical rules of historical data. Step B: Establishing an energy storage and load model, the energy storage and load model includes: an energy storage model and a load absorption energy consumption model; the energy storage model realizes peak shaving and valley filling while suppressing the anti-load characteristics of new energy power generation; the load absorption energy consumption model adjusts the operation time of the equipment according to the photovoltaic output characteristics, transfers the time-shiftable load during the peak power consumption period to the low power consumption period or the peak period of photovoltaic output, and performs energy consumption analysis according to the proportion of different load categories; Step C: Establish an optimization model for the photovoltaic microgrid system with energy storage system and demand response, measure the photovoltaic absorption capacity of the distribution network by the sum of the distributed photovoltaic access capacity of each node, and use the random weighting method to weight the two objective functions of maximum photovoltaic absorption capacity and minimum active network loss, and integrate them into the optimization goal. Combined with the power flow constraints, node voltage deviation constraints, distributed photovoltaic power generation energy constraints and energy storage system energy upper limit constraints, the photovoltaic absorption capacity and economy of the distribution network are considered to achieve the optimization of the distributed photovoltaic microgrid system.

2. The distributed photovoltaic microgrid system optimization method considering demand-side response and energy storage according to claim 1 is characterized in that: The step A of establishing a photovoltaic day-ahead output prediction model includes the following steps: Step A1: Based on the least squares support vector machine, the inequality constraints in the optimization formula are converted into equality constraints; based on the data sample set, the least squares support vector machine regression algorithm is used, and the nonlinear mapping φ(x) is used to project the input data into a high-dimensional feature space, so that the function value y corresponding to the sample x is approximately expressed as f(x); y=f(x)=ωφ(x)+b (1) In the formula, ω is the weight vector in the feature space, and b is the bias; The data sample set is: D = {(x i ,y i )|(i=1,2,...,l)}, where: D represents the data sample set, l is the number of samples in the sample set, x i is the i-th sample data, y i is the function value corresponding to the i-th sample data; The loss function of the photovoltaic day-ahead output prediction model is defined as: In the formula, ε is the insensitive coefficient, e is the loss function, f(x i ) is the function value corresponding to the i-th sample; The least squares support vector regression is expressed as an optimization problem for equation (2): In the formula, γ is the penalty factor, J(ω,e) is the least squares support vector, φ(x i ) is x i The nonlinear mapping, e i is the sample error calculated by formula (2), y i Represents the function value corresponding to the i-th sample data; The Lagrangian function is established based on formula (3), namely: In the formula, a i is the Lagrange multiplier, L is the Lagrange function, a represents the set of Lagrange multipliers, e i is the sample error calculated by formula (2); According to the KKT condition, the linear equation is: Where I is the l×l unit matrix, γ -1 is the reciprocal of the penalty factor, b is the bias, and the remaining expressions are: In the formula, e l is the identity matrix, Ω i,j For x i With x j The product of the nonlinear mapping, a is the Lagrange multiplier matrix, y is the sample function value matrix, K(x i ,x j ) is the kernel function, T represents transpose; Solve the linear equations in equation (6) to obtain the values ​​of a and b, and obtain the regression fitting function, which is: In the formula, K(x,x i ) is a kernel function, and the kernel function adopts a radial basis kernel function, and the radial basis kernel function is: Where σ is the width of the radial basis kernel function; Step A2: performing optimization based on a particle swarm optimization algorithm and outputting an optimization result, wherein the optimization result includes: an optimal penalty parameter c and an optimal kernel function parameter g; Suppose in a D-dimensional search space, there are n particles forming a population M = (M1, M2, ..., M n ), where the position of the i-th particle is X i =(X i1 ,X i2 ,...,X iD ) T , speed is V i =(V i1 ,V i2 ,...,V iD ) T , the individual extreme value is P i =(P i1 ,P i2 ,...,P iD ) T , the global extreme value of the population is P g =(P g1 ,P g2 ,...,P gD ) T ; The evolution mode of the particle swarm optimization algorithm is: Particle swarm algorithm evolution model = particle historical inertia + particle self-cognition + particle group cognition The update formula of velocity and position in the particle iterative optimization process is expressed as: V id (k+1)=ωV id (k)+c1r1[P id (k)-X id (k)]+c2r2[P gd (k)-X id (k)] (9) X id (k+1)=X id (k)+V id (k+1) (10) Where V id (k) is the optimization speed of the current particle, P id (k) is the optimal position of the current particle, X id (k) Current particle position, P gd (k) is the optimal position of the current particle group, ω is the inertia factor, which is mainly used to balance the global search and local search capabilities; d = 1, 2, ..., n, d represents the number of iterations; i = 1, 2, ..., D, i represents the i-th particle; k is the evolutionary generation; c1 and c2 are acceleration factors, which are non-negative values, c1 represents the degree of inheritance of the particle to its own optimal position, and c2 represents the degree of inheritance of the particle to the global optimal position; r1 and r2 are random numbers, and the value space is [0, 1]; Each particle determines the adaptive value based on the objective function during each iteration and determines the optimal position P of the current particle. id (k) and the optimal position of the group P gd (k), update the speed and position of each particle through equation (9) and equation (10); when the number of iterations reaches the set value or the result meets the accuracy requirement, the optimization is terminated, otherwise the iteration continues.

3. The distributed photovoltaic microgrid system optimization method considering demand-side response and energy storage according to claim 1 is characterized in that: The photovoltaic power generation prediction method based on the particle swarm algorithm to improve the support vector machine in step A includes the following steps: Import historical photovoltaic data; Use particle swarm algorithm for optimization; Determine whether the optimization conditions are met; If the optimization condition is not met, update the historical optimal position of each particle; update the global optimal position of the group; If the optimization condition is met, the optimization result is output, and the optimization result includes: the best penalty parameter c and the best kernel function parameter g; Import the optimization results into support vector regression for training and prediction; Calculate mean square error and coefficient of determination; Output predicted power.

4. The distributed photovoltaic microgrid system optimization method considering demand-side response and energy storage according to claim 1 is characterized in that: The step B of establishing the energy storage and load model comprises the following steps: Step B1: Said The energy storage model is based on the power supply characteristics and load characteristics of the energy storage system. It can store new energy electricity in the form of load during the off-peak period; during the peak period, it releases the electricity stored during the off-peak period to relieve the power supply pressure of the power system; charging and discharging cannot be carried out at the same time in the same period, and the charging power and discharge power The following constraints should be met at time t: When the distributed photovoltaic output is greater than the load, When the distributed photovoltaic output is less than the load, Step B2: The load absorption energy consumption model calculates the load absorption of photovoltaic power in the domain. By coordinating the time-shiftable load and the fixed load, the energy consumption analysis is performed according to the proportion of different load categories. The energy consumption calculation method of load absorption is as follows: C ec =p c ∫t c L c (t)dt+E c (12) In the formula, C ec is the energy consumption of load absorption, p c It is the time-of-use electricity price, which includes the electricity prices in peak, flat and valley periods; t c is the load operation time; L c is the load power; E c For running costs.

5. The distributed photovoltaic microgrid system optimization method considering demand-side response and energy storage according to claim 1 is characterized in that: The calculation method of the photovoltaic absorption capacity in step C is: Where f1 is the photovoltaic absorption capacity; N PV P is the total number of nodes connected to the photovoltaic system; PV,m is the photovoltaic access capacity of the mth node.

6. The distributed photovoltaic microgrid system optimization method considering demand-side response and energy storage according to claim 1 is characterized in that: The calculation method of active network loss in step C is: Where f2 is the active network loss, {N B } is the set of all nodes in the system; P ij ,Q ij are respectively the active power and reactive power flowing through branch ij; U i is the voltage value of node i; R ij is the line resistance between nodes i and j.

7. The distributed photovoltaic microgrid system optimization method considering demand-side response and energy storage according to claim 1 is characterized in that: The function expression of the random weighting method in step C is: max f=ω1f1-ω2f2 (15) Where f is the total objective function, ω1, ω2 are random numbers, and ω1+ω2=1, ω1≤1, ω2≤1.

Citation Information

Patent Citations

  • A grid-connected micro-grid economic optimization scheduling strategy for promoting photovoltaic consumption and considering combined heat and power generation and demand response

    CN109816556A

  • Wind storage combined heat supply system optimization scheduling strategy considering electric heat demand response

    CN113852139A

  • Source-load integrated prediction method based on regression analysis and LSSVM

    CN115423143A

  • Power distribution network demand side resource adjustment demand quantification method, system and equipment

    CN116914798A

  • Microgrid voltage control method considering price type demand response and inverter droop parameter optimization

    CN117578491A

Cited By

  • Photovoltaic dynamic consumption optimization method, device and equipment for expressway service area and storage medium

    CN120764794A

  • Intelligent evaluation method and system for distributed photovoltaic resources

    CN121124201A