An AVC control method for photovoltaic power station based on particle swarm algorithm

Abstract

The present application relates to a kind of photovoltaic power station AVC control method based on particle swarm algorithm, belong to new energy power station monitoring automation technical field.The method includes: through photovoltaic power station automatic voltage control AVC system setting main transformer high voltage side bus voltage U pcc_ref With the minimum deviation of photovoltaic power station main transformer high voltage side bus voltage and the minimum station active loss as target, a multi-objective optimization model is established;Using fuzzy theory, the multi-objective model is converted into a single objective model;Particle swarm optimization algorithm is used to solve the single objective model;According to the solution, the corresponding reactive power regulation instructions are issued by the AVC system to each reactive power source.The method of the present application considers the active loss in the station when distributing reactive power in the AVC system of photovoltaic power station, which can effectively reduce the active loss in the station while meeting the AVC voltage control, and provides a strong support for safe and economic operation of power station.

Description

Technical Field

[0001] This invention belongs to the field of new energy power plant monitoring automation technology, specifically relating to an AVC control method for photovoltaic power plants based on particle swarm optimization algorithm. Background Technology

[0002] In recent years, with the increasing maturity of photovoltaic power generation technology, a large number of photovoltaic power plants have been connected to the grid. The active power output of photovoltaic power plants is random and fluctuating, which has a certain impact on their reactive power distribution, active power loss, and the voltage of the high-voltage busbar of the main transformer. Therefore, photovoltaic power plants must have a certain reactive power and voltage control capability to facilitate the safe and economical operation of the power grid and the photovoltaic power plants themselves.

[0003] To improve power quality, automatic voltage control (AVC) systems are required in photovoltaic (PV) power plants. Currently, the main reactive power distribution strategies for PV power plants include distribution based on regulation capacity, distribution based on regulation margin, equal power factor distribution, and average distribution. However, none of these strategies consider active power losses during reactive power distribution. Therefore, researching an AVC control method that considers active power losses in PV power plants during reactive power distribution has significant engineering practical value. Summary of the Invention

[0004] Currently, photovoltaic power plant AVC systems do not consider active power losses within the plant when allocating reactive power, resulting in high active power losses within the plant itself. The purpose of this invention is to address the shortcomings of existing technologies by providing a photovoltaic power plant AVC control method based on particle swarm optimization (PSO). This method aims to minimize the voltage deviation of the high-voltage side bus of the main transformer and the active power losses within the plant. The optimal solution, i.e., the optimal reactive power allocation strategy, is obtained through PSO optimization. This method can effectively reduce active power losses within the plant while satisfying AVC voltage control requirements, and can effectively reduce voltage fluctuations on the high-voltage side bus of the main transformer and active power losses within the photovoltaic power plant.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A photovoltaic power plant AVC control method based on particle swarm optimization algorithm includes the following steps:

[0007] The voltage U of the high-voltage busbar of the main transformer is set through the automatic voltage control (AVC) system of the photovoltaic power station. pcc_ref ;

[0008] A multi-objective optimization model is established with the objectives of minimizing the voltage deviation of the high-voltage side busbar of the main transformer in the photovoltaic power station and minimizing the active power loss within the station.

[0009] Using fuzzy theory, a multi-objective model is transformed into a single-objective model;

[0010] The particle swarm optimization algorithm is used to optimize and solve the single-objective model;

[0011] Based on the solution results, the AVC system issues corresponding reactive power adjustment commands to each reactive power source.

[0012] Furthermore, preferably, the multi-objective optimization model includes an objective function that aims to minimize the active power loss within the photovoltaic power station, an objective function that aims to minimize the voltage deviation of the high-voltage side bus of the main transformer of the photovoltaic power station, and constraints; the constraints include power flow equation constraints, control variable constraints, and state variable constraints.

[0013] Furthermore, preferably, the specific method for establishing a multi-objective optimization model is as follows:

[0014] (1) With the goal of minimizing the active power loss within the photovoltaic power station, the objective function is expressed as:

[0015] (1)

[0016] (2)

[0017] (3)

[0018] In the formula: U pcc U is the voltage of the high-voltage side busbar of the main transformer in the photovoltaic power station. pcc_ref This indicates the setpoint of the high-voltage bus voltage on the main transformer side of the photovoltaic power station; U pcc1 The voltage of the low-voltage busbar of the main transformer of the photovoltaic power station is given by: k = number of photovoltaic power generation units; l = number of collector lines; n = number of SVGs; P loss_T P loss_L P represents the active power loss of the power station's main transformer and the transmission line, respectively; loss_t P loss_l These represent the active power losses on the connecting transformer and the collector line, respectively; R T R t R L R l and R SVG_η Let P and Q represent the resistances of the main transformer, connecting transformer, transmission line, collector line, and the ηth SVG (η=1,2,...n), respectively; P and Q represent the active power and reactive power generated by the photovoltaic power station, respectively; P i_j Q i_j U represents the active power and reactive power generated by the j-th photovoltaic power generation unit on the i-th collector line, respectively; i_j Q is the output voltage of the j-th photovoltaic power generation unit on the i-th collector line; SVG_η For the reactive power output of the ηth SVG; P i_m Q i_m Let them represent the active power and reactive power generated by the m-th photovoltaic power generation unit on the i-th collector line, respectively;

[0019] (2) Taking the minimum voltage deviation of the high-voltage side busbar of the main transformer of the photovoltaic power station as the objective, the objective function is expressed as:

[0020] (4)

[0021] In the formula: ΔU pcc_max This indicates the maximum allowable deviation of the voltage on the high-voltage side busbar of the main transformer in a photovoltaic power station;

[0022] (3) The power flow equation constraints are:

[0023] (5)

[0024] In the formula: P α Q α Let g represent the active and reactive power injected into node α, respectively; α∈β, representing all nodes connected to node α; g αβ and b αβ U represents the line conductance and susceptance connecting nodes α and β, respectively; α U β θ represents the voltages at nodes α and β, respectively; αβ This represents the phase angle difference between the voltages at nodes α and β.

[0025] (4) The control variable constraints are:

[0026] (6)

[0027] In the formula: Q i_jmax and Q i_jmin Q represents the maximum and minimum reactive power output of the j-th photovoltaic power generation unit on the i-th collector line, respectively; SVG_ηmax and Q SVG_ηmin Representing the ηth The maximum and minimum reactive power output of the SVG;

[0028] (5) The state variable constraints are:

[0029] (7)

[0030] In the formula: U pcc_max U pcc_min These represent the maximum and minimum allowable values ​​of the high-voltage bus voltage of the main transformer, respectively.

[0031] Furthermore, preferably, the specific method for transforming a multi-objective model into a single-objective model using fuzzy theory is as follows:

[0032] A single compromise model is established by scaling the active power loss and the voltage deviation of the high-voltage side bus of the main transformer as follows:

[0033] (8)

[0034] In the formula: μ1(F1) and μ2(F2) represent the membership values ​​of active power loss and voltage deviation, respectively; μ1 and μ2 are weighting coefficients, and μ1+μ2=1. When calculating formula (8), if you want to focus on active power loss, you can set a larger μ1, and vice versa. The power station can set the values ​​of μ1 and μ2 according to the actual situation; both active power loss and voltage deviation are minimum objective functions. According to fuzzy theory, the membership functions of μ1(F1) and μ2(F2) are constructed as follows:

[0035] (9)

[0036] (10)

[0037] In the formula: F 1min F 1max These represent the minimum and maximum active power losses of the power station, respectively; F 2min F 2max These represent the minimum and maximum set voltage deviations, respectively.

[0038] Furthermore, preferably, the specific method for optimizing and solving a single-objective model using the particle swarm optimization algorithm is as follows:

[0039] (1) Determine the algorithm parameters; including population size N, maximum number of iterations T max Particle dimension D, inertia weight ω, and learning factors c1 and c2;

[0040] (2) Determine the structure X of the particles and initialize the particle swarm;

[0041] (12)

[0042] In the formula: M is the number of photovoltaic inverters, Q M This represents the reactive power output of the Mth photovoltaic inverter;

[0043] (3) Use the Newton-Raphson method to calculate the power flow. Calculate the fitness value minF of each particle according to formula (8). Set the current position of each particle as the individual extreme value. Then evaluate all particles and select the individual with the smallest fitness value as the population extreme value.

[0044] (4) Update the velocity and position of all particles according to equation (11);

[0045] At the (t+1)th iteration, the velocity v of the λ-th particle in d dimension λd and position x λd The update formula is:

[0046] (11)

[0047] In the formula: λ=1.2,...,N, where N is the sample population size; d=1.2,...,D, where D is the dimension of the search space; t is the number of iterations; ω is the inertia weight; r1 and r2 are random numbers between [0,1]; c1 and c2 are learning factors; P λd (t) represents the optimal position experienced by the λ-th particle in the d-th iteration; P gd (t) represents the optimal position experienced by all particles in d dimensions at the t-th iteration;

[0048] v λd (t) represents the d-dimensional velocity of the λ-th particle at the t-th iteration;

[0049] x λd (t) represents the d-dimensional position of the λ-th particle at the t-th iteration;

[0050] (5) Recalculate the fitness values ​​of all particles, evaluate them, and update the individual extreme values ​​and the population extreme values;

[0051] (6) Determine whether t ≥ T max If yes, output the result; otherwise, return to step (4).

[0052] In this invention, when calculating equation (8), if it is desired to prioritize active power loss, a larger μ1 can be set; otherwise, a larger μ2 can be set. The power plant can set the values ​​of μ1 and μ2 according to the actual situation. This invention does not impose any limitations on this.

[0053] According to the particle swarm algorithm, the final result of equation (12) is denoted as X0. The AVC system issues corresponding reactive power adjustment commands to each reactive power source.

[0054] In this invention, P λd (t) represents the optimal position experienced by the λ-th particle in the d-th iteration, which is the individual extreme value; P gd (t) represents the optimal position experienced by all particles in d dimension at the t-th iteration, which is the population extreme value.

[0055] The Newton-Raphson method used in this invention can be found in Chapter 4, Section 3 of "Steady-State Analysis of Power Systems (3rd Edition)" published by China Electric Power Press.

[0056] Compared with the prior art, the beneficial effects of this invention are as follows:

[0057] The method of this invention aims to minimize the voltage deviation of the high-voltage side busbar of the main transformer in the power plant and the active power loss within the plant. It uses the particle swarm optimization algorithm to obtain the optimal solution, namely the optimal reactive power allocation strategy. This method can effectively reduce the active power loss within the plant while satisfying AVC voltage control, and can effectively reduce the voltage fluctuation of the high-voltage side busbar of the main transformer in the photovoltaic power plant and the active power loss within the plant.

[0058] Traditional AVC control methods do not consider active power losses within the station when distributing reactive power. As can be seen from application examples, the method of this invention significantly reduces active power losses within the station compared to the traditional method. Therefore, using this method can reduce the active power losses of photovoltaic power plants to a certain extent while ensuring that the bus voltage regulation is in place. Compared with the traditional AVC control method, it improves the economic efficiency of power plant operation and is easy to promote and apply. Attached Figure Description

[0059] Figure 1 This is a flowchart of the AVC control method for photovoltaic power plants based on particle swarm optimization algorithm according to the present invention.

[0060] Figure 2 This is the convergence curve of the fitness value of the objective function in the application example. Detailed Implementation

[0061] The present invention will now be described in further detail with reference to the embodiments.

[0062] Those skilled in the art will understand that the following embodiments are for illustrative purposes only and should not be construed as limiting the scope of the invention. Where specific techniques or conditions are not specified in the embodiments, they are performed in accordance with the techniques or conditions described in the literature in the field or according to the product instructions. Materials or equipment whose manufacturers are not specified are all conventional products that can be obtained by purchase.

[0063] Example 1

[0064] A photovoltaic power plant AVC control method based on particle swarm optimization algorithm includes the following steps:

[0065] The voltage U of the high-voltage busbar of the main transformer is set through the automatic voltage control (AVC) system of the photovoltaic power station. pcc_ref ;

[0066] A multi-objective optimization model is established with the objectives of minimizing the voltage deviation of the high-voltage side busbar of the main transformer in the photovoltaic power station and minimizing the active power loss within the station.

[0067] Using fuzzy theory, a multi-objective model is transformed into a single-objective model;

[0068] The particle swarm optimization algorithm is used to optimize and solve the single-objective model;

[0069] Based on the solution results, the AVC system issues corresponding reactive power adjustment commands to each reactive power source.

[0070] Example 2

[0071] A photovoltaic power plant AVC control method based on particle swarm optimization algorithm includes the following steps:

[0072] The voltage U of the high-voltage busbar of the main transformer is set through the automatic voltage control (AVC) system of the photovoltaic power station. pcc_ref ;

[0073] A multi-objective optimization model is established with the objectives of minimizing the voltage deviation of the high-voltage side busbar of the main transformer in the photovoltaic power station and minimizing the active power loss within the station.

[0074] Using fuzzy theory, a multi-objective model is transformed into a single-objective model;

[0075] The particle swarm optimization algorithm is used to optimize and solve the single-objective model;

[0076] Based on the solution results, the AVC system issues corresponding reactive power adjustment commands to each reactive power source.

[0077] The multi-objective optimization model includes an objective function that aims to minimize the active power loss within the photovoltaic power station, an objective function that aims to minimize the voltage deviation of the high-voltage side bus of the main transformer of the photovoltaic power station, and constraints; the constraints include power flow equation constraints, control variable constraints, and state variable constraints.

[0078] The specific method for establishing a multi-objective optimization model is as follows:

[0079] (1) With the goal of minimizing the active power loss within the photovoltaic power station, the objective function is expressed as:

[0080] (1)

[0081] (2)

[0082] (3)

[0083] In the formula: U pcc U is the voltage of the high-voltage side busbar of the main transformer in the photovoltaic power station. pcc_ref This indicates the setpoint of the high-voltage bus voltage on the main transformer side of the photovoltaic power station; U pcc1 The voltage of the low-voltage busbar of the main transformer of the photovoltaic power station is given by: k = number of photovoltaic power generation units; l = number of collector lines; n = number of SVGs; P loss_T P loss_L P represents the active power loss of the power station's main transformer and the transmission line, respectively; loss_t P loss_l These represent the active power losses on the connecting transformer and the collector line, respectively; R T R t RL R l and R SVG_η Let P and Q represent the resistances of the main transformer, connecting transformer, transmission line, collector line, and the ηth SVG (η=1,2,...n), respectively; P and Q represent the active power and reactive power generated by the photovoltaic power station, respectively; Q SVG For the reactive power output of SVG; P i_j Q i_j U represents the active power and reactive power generated by the j-th photovoltaic power generation unit on the i-th collector line, respectively; i_j Q is the output voltage of the j-th photovoltaic power generation unit on the i-th collector line; SVG_η For the reactive power output of the ηth SVG; P i_m Q i_m Let them represent the active power and reactive power generated by the m-th photovoltaic power generation unit on the i-th collector line, respectively;

[0084] (2) Taking the minimum voltage deviation of the high-voltage side busbar of the main transformer of the photovoltaic power station as the objective, the objective function is expressed as:

[0085] (4)

[0086] In the formula: ΔU pcc_max This indicates the maximum allowable deviation of the voltage on the high-voltage side busbar of the main transformer in a photovoltaic power station;

[0087] (3) The power flow equation constraints are:

[0088] (5)

[0089] In the formula: P α Q α Let g represent the active and reactive power injected into node α, respectively; α∈β, representing all nodes connected to node α; g αβ and b αβ U represents the line conductance and susceptance connecting nodes α and β, respectively; α U β θ represents the voltages at nodes α and β, respectively; αβ This represents the phase angle difference between the voltages at nodes α and β.

[0090] (4) The control variable constraints are:

[0091] (6)

[0092] In the formula: Q i_jmax and Q i_jmin Q represents the maximum and minimum reactive power output of the j-th photovoltaic power generation unit on the i-th collector line, respectively; SVG_ηmax and Q SVG_ηmin Representing the ηth The maximum and minimum reactive power output of the SVG;

[0093] (5) The state variable constraints are:

[0094] (7)

[0095] In the formula: U pcc_max U pcc_min These represent the maximum and minimum allowable values ​​of the high-voltage bus voltage of the main transformer, respectively.

[0096] The specific method for transforming a multi-objective model into a single-objective model using fuzzy theory is as follows:

[0097] A single compromise model is established by scaling the active power loss and the voltage deviation of the high-voltage side bus of the main transformer as follows:

[0098] (8)

[0099] In the formula: μ1(F1) and μ2(F2) represent the membership values ​​of active power loss and voltage deviation, respectively; μ1 and μ2 are weighting coefficients, and μ1+μ2=1. When calculating formula (8), if you want to focus on active power loss, you can set a larger μ1, and vice versa. The power station can set the values ​​of μ1 and μ2 according to the actual situation; both active power loss and voltage deviation are minimum objective functions. According to fuzzy theory, the membership functions of μ1(F1) and μ2(F2) are constructed as follows:

[0100] (9)

[0101] (10)

[0102] In the formula: F 1min F 1max These represent the minimum and maximum active power losses of the power station, respectively; F 2min F 2max These represent the minimum and maximum set voltage deviations, respectively.

[0103] Example 3

[0104] A photovoltaic power plant AVC control method based on particle swarm optimization algorithm includes the following steps:

[0105] The voltage U of the high-voltage busbar of the main transformer is set through the automatic voltage control (AVC) system of the photovoltaic power station. pcc_ref ;

[0106] A multi-objective optimization model is established with the objectives of minimizing the voltage deviation of the high-voltage side busbar of the main transformer in the photovoltaic power station and minimizing the active power loss within the station.

[0107] Using fuzzy theory, a multi-objective model is transformed into a single-objective model;

[0108] The particle swarm optimization algorithm is used to optimize and solve the single-objective model;

[0109] Based on the solution results, the AVC system issues corresponding reactive power adjustment commands to each reactive power source.

[0110] The multi-objective optimization model includes an objective function that aims to minimize the active power loss within the photovoltaic power station, an objective function that aims to minimize the voltage deviation of the high-voltage side bus of the main transformer of the photovoltaic power station, and constraints; the constraints include power flow equation constraints, control variable constraints, and state variable constraints.

[0111] The specific method for establishing a multi-objective optimization model is as follows:

[0112] (1) With the goal of minimizing the active power loss within the photovoltaic power station, the objective function is expressed as:

[0113] (1)

[0114] (2)

[0115] (3)

[0116] In the formula: U pcc U is the voltage of the high-voltage side busbar of the main transformer in the photovoltaic power station. pcc_ref This indicates the setpoint of the high-voltage bus voltage on the main transformer side of the photovoltaic power station; U pcc1 The voltage of the low-voltage busbar of the main transformer of the photovoltaic power station is given by: k = number of photovoltaic power generation units; l = number of collector lines; n = number of SVGs; P loss_T P loss_L P represents the active power loss of the power station's main transformer and the transmission line, respectively; loss_t P loss_l These represent the active power losses on the connecting transformer and the collector line, respectively; R T R t R L R l and R SVG_η Let P and Q represent the resistances of the main transformer, connecting transformer, transmission line, collector line, and the ηth SVG (η=1,2,...n), respectively; P and Q represent the active power and reactive power generated by the photovoltaic power station, respectively; P i_j Q i_j U represents the active power and reactive power generated by the j-th photovoltaic power generation unit on the i-th collector line, respectively; i_j Q is the output voltage of the j-th photovoltaic power generation unit on the i-th collector line; SVG_η For the reactive power output of the ηth SVG; Pi_m Q i_m Let them represent the active power and reactive power generated by the m-th photovoltaic power generation unit on the i-th collector line, respectively;

[0117] (2) Taking the minimum voltage deviation of the high-voltage side busbar of the main transformer of the photovoltaic power station as the objective, the objective function is expressed as:

[0118] (4)

[0119] In the formula: ΔU pcc_max This indicates the maximum allowable deviation of the voltage on the high-voltage side busbar of the main transformer in a photovoltaic power station;

[0120] (3) The power flow equation constraints are:

[0121] (5)

[0122] In the formula: P α Q α Let g represent the active and reactive power injected into node α, respectively; α∈β, representing all nodes connected to node α; g αβ and b αβ U represents the line conductance and susceptance connecting nodes α and β, respectively; α U β θ represents the voltages at nodes α and β, respectively; αβ This represents the phase angle difference between the voltages at nodes α and β.

[0123] (4) The control variable constraints are:

[0124] (6)

[0125] In the formula: Q i_jmax and Q i_jmin Q represents the maximum and minimum reactive power output of the j-th photovoltaic power generation unit on the i-th collector line, respectively; SVG_ηmax and Q SVG_ηmin Representing the ηth The maximum and minimum reactive power output of the SVG;

[0126] (5) The state variable constraints are:

[0127] (7)

[0128] In the formula: U pcc_max U pcc_min These represent the maximum and minimum allowable values ​​of the high-voltage bus voltage of the main transformer, respectively.

[0129] The specific method for transforming a multi-objective model into a single-objective model using fuzzy theory is as follows:

[0130] A single compromise model is established by scaling the active power loss and the voltage deviation of the high-voltage side bus of the main transformer as follows:

[0131] (8)

[0132] In the formula: μ1(F1) and μ2(F2) represent the membership values ​​of active power loss and voltage deviation, respectively; μ1 and μ2 are weighting coefficients, and μ1 + μ2 = 1; both active power loss and voltage deviation are minimum objective functions. According to fuzzy theory, the membership functions for μ1(F1) and μ2(F2) are constructed as follows:

[0133] (9)

[0134] (10)

[0135] In the formula: F 1min F 1max These represent the minimum and maximum active power losses of the power station, respectively; F 2min F 2max These represent the minimum and maximum set voltage deviations, respectively.

[0136] The specific method for optimizing a single-objective model using the particle swarm optimization algorithm is as follows:

[0137] (1) Determine the algorithm parameters; including population size N, maximum number of iterations T max Particle dimension D, inertia weight ω, and learning factors c1 and c2;

[0138] (2) Determine the structure X of the particles and initialize the particle swarm;

[0139] (12)

[0140] In the formula: M is the number of photovoltaic inverters, Q M This represents the reactive power output of the Mth photovoltaic inverter;

[0141] (3) Use the Newton-Raphson method to calculate the power flow. Calculate the fitness value minF of each particle according to formula (8). Set the current position of each particle as the individual extreme value. Then evaluate all particles and select the individual with the smallest fitness value as the population extreme value.

[0142] (4) Update the velocity and position of all particles according to equation (11);

[0143] At the (t+1)th iteration, the velocity v of the λ-th particle in d dimension λd and position x λd The update formula is:

[0144] (11)

[0145] In the formula: λ=1.2,...,N, where N is the sample population size; d=1.2,...,D, where D is the dimension of the search space; t is the number of iterations; ω is the inertia weight; r1 and r2 are random numbers between [0,1]; c1 and c2 are learning factors; P λd (t) represents the optimal position experienced by the λ-th particle in the d-th iteration; P gd (t) represents the optimal position experienced by all particles in d dimensions at the t-th iteration;

[0146] v λd (t) represents the d-dimensional velocity of the λ-th particle at the t-th iteration;

[0147] x λd (t) represents the d-dimensional position of the λ-th particle at the t-th iteration;

[0148] (5) Recalculate the fitness values ​​of all particles, evaluate them, and update the individual extreme values ​​and the population extreme values;

[0149] (6) Determine whether t ≥ T max If yes, output the result; otherwise, return to step (4).

[0150] Application Examples

[0151] A photovoltaic power plant AVC control method based on particle swarm optimization algorithm, such as Figure 1 As shown, it includes the following steps:

[0152] (i) Setting the main transformer high-voltage side bus voltage U through the photovoltaic power station automatic voltage control (AVC) system. pcc_ref ;

[0153] (ii) A multi-objective optimization model is established with the objectives of minimizing the voltage deviation of the high-voltage side busbar of the main transformer of the photovoltaic power station and minimizing the active power loss within the station;

[0154] (1) With the goal of minimizing the active power loss within the photovoltaic power station, the objective function is expressed as:

[0155] (1)

[0156] (2)

[0157] (3)

[0158] In the formula: U pcc U is the voltage of the high-voltage side busbar of the main transformer in the photovoltaic power station. pcc_ref This indicates the setpoint of the high-voltage bus voltage on the main transformer side of the photovoltaic power station; U pcc1The voltage of the low-voltage busbar of the main transformer of the photovoltaic power station is given by: k = number of photovoltaic power generation units; l = number of collector lines; n = number of SVGs; P loss_T P loss_L P represents the active power loss of the power station's main transformer and the transmission line, respectively; loss_t P loss_l These represent the active power losses on the connecting transformer and the collector line, respectively; R T R t R L R l and R SVG_η Let P and Q represent the resistances of the main transformer, connecting transformer, transmission line, collector line, and the ηth SVG (η=1,2,...n), respectively; P and Q represent the active power and reactive power generated by the photovoltaic power station, respectively; P i_j Q i_j U represents the active power and reactive power generated by the j-th photovoltaic power generation unit on the i-th collector line, respectively; i_j Q is the output voltage of the j-th photovoltaic power generation unit on the i-th collector line; SVG_η For the reactive power output of the ηth SVG; P i_m Q i_m Let them represent the active power and reactive power generated by the m-th photovoltaic power generation unit on the i-th collector line, respectively;

[0159] (2) Taking the minimum voltage deviation of the high-voltage side busbar of the main transformer of the photovoltaic power station as the objective, the objective function is expressed as:

[0160] (4)

[0161] In the formula: ΔU pcc_max This indicates the maximum allowable deviation of the voltage on the high-voltage side busbar of the main transformer in a photovoltaic power station;

[0162] (3) The power flow equation constraints are:

[0163] (5)

[0164] In the formula: P α Q α Let g represent the active and reactive power injected into node α, respectively; α∈β, representing all nodes connected to node α; g αβ and b αβ U represents the line conductance and susceptance connecting nodes α and β, respectively; α U β θ represents the voltages at nodes α and β, respectively; αβ This represents the phase angle difference between the voltages at nodes α and β.

[0165] (4) The control variable constraints are:

[0166] (6)

[0167] In the formula: Q i_jmax and Q i_jmin Q represents the maximum and minimum reactive power output of the j-th photovoltaic power generation unit on the i-th collector line, respectively; SVG_ηmax and Q SVG_ηmin These represent the maximum and minimum values ​​of the reactive power output of the ηth SVG, respectively;

[0168] (5) The state variable constraints are:

[0169] (7)

[0170] In the formula: U pcc_max U pcc_min These represent the maximum and minimum allowable values ​​of the high-voltage bus voltage of the main transformer, respectively.

[0171] (iii) Using fuzzy theory to transform multi-objective models into single-objective models;

[0172] A single compromise model is established by scaling the active power loss and the voltage deviation of the high-voltage side bus of the main transformer as follows:

[0173] (8)

[0174] In the formula: μ1(F1) and μ2(F2) represent the membership values ​​of active power loss and voltage deviation, respectively; μ1 and μ2 are weighting coefficients, and μ1+μ2=1. When calculating formula (8), if you want to focus on active power loss, you can set a larger μ1, and vice versa. The power station can set the values ​​of μ1 and μ2 according to the actual situation; for example, set μ1=0.3, μ2=7, or μ1=0.6, μ2=0.4, etc. This invention does not limit this.

[0175] Both active power loss and voltage deviation are minimum objective functions. Based on fuzzy theory, membership functions are constructed for μ1(F1) and μ2(F2) respectively as follows:

[0176] (9)

[0177] (10)

[0178] In the formula: F 1min F 1max These represent the minimum and maximum active power losses of the power station, respectively; F 2min F 2max These represent the minimum and maximum set voltage deviations, respectively.

[0179] (iv) The particle swarm optimization algorithm is used to optimize and solve equation (8);

[0180] (1) Determine the algorithm parameters; including population size N, maximum number of iterations T max Particle dimension D, inertia weight ω, and learning factors c1 and c2;

[0181] (2) Determine the structure X of the particles and initialize the particle swarm;

[0182] (12)

[0183] In the formula: M is the number of photovoltaic inverters, Q M This represents the reactive power output of the Mth photovoltaic inverter;

[0184] (3) Use the Newton-Raphson method to calculate the power flow. Calculate the fitness value minF of each particle according to formula (8). Set the current position of each particle as the individual extreme value. Then evaluate all particles and select the individual with the smallest fitness value as the population extreme value.

[0185] (4) Update the velocity and position of all particles according to equation (11);

[0186] At the (t+1)th iteration, the velocity v of the λ-th particle in d dimension λd and position x λd The update formula is:

[0187] (11)

[0188] In the formula: λ=1.2,...,N, where N is the sample population size; d=1.2,...,D, where D is the dimension of the search space; t is the number of iterations; ω is the inertia weight; r1 and r2 are random numbers between [0,1]; c1 and c2 are learning factors; P λd (t) represents the optimal position experienced by the λ-th particle in the d-th iteration; P gd (t) represents the optimal position experienced by all particles in d dimensions at the t-th iteration;

[0189] v λd (t) represents the d-dimensional velocity of the λ-th particle at the t-th iteration;

[0190] x λd (t) represents the d-dimensional position of the λ-th particle at the t-th iteration;

[0191] (5) Recalculate the fitness values ​​of all particles, evaluate them, and update the individual extreme values ​​and the population extreme values;

[0192] (6) Determine whether t ≥ T maxIf yes, output the result; otherwise, return to step (4).

[0193] (v) Based on the final result X0 of equation (12) obtained by the particle swarm algorithm, the AVC system issues a reactive power adjustment command to each reactive power source.

[0194] Taking a photovoltaic power station as an example, a simulation was performed using Matlab. The power station has an installed capacity of 40MW and consists of four collector lines. Each collector line connects 10 sets of photovoltaic power generation units in parallel. Each photovoltaic power generation unit consists of a 1MW inverter and several photovoltaic arrays. The units are fed into the collector line via a 0.29 / 10kV step-up transformer. The distance between adjacent photovoltaic power generation units is 1km. The collector lines connecting the photovoltaic power generation units use YJV23-8.7 / 10, 3×150 mm... 2 The cable connections are as follows; the inverter's reactive power output range is -478KVar (capacitive) to 478KVar (inductive); the main transformer capacity is 60MVA, with a transformation ratio of 121kV / 10.5kV; the transmission line is an LGJ-400 overhead line with a length of 40km; the power station is equipped with an SVG with a capacity of ±6Mvar; the maximum number of iterations of the particle swarm optimization algorithm is set to 50, the particle population size is 30, ω=0.8, c1=c2=2; F 1max =10MW, F 1min =0MW;F 2max =2,F 2min =0. Note: Voltage deviation has no unit.

[0195] The AVC system sets the 110kV bus voltage of the substation to 115kV, with an AVC control dead zone of ±200V and ΔU. pcc_max =200V; U pcc_max =115.2kV, U pcc_min =114.8kV; the active power of the substation is 36.089MW at this time; Scheme 1 sets μ1=0.15, μ2=0.85; the traditional AVC control method does not consider the active power loss in the substation when performing reactive power distribution. For comparative analysis, Scheme 2 sets μ1=0, μ2=1, that is, it does not consider the loss in the substation when calculating the output of each reactive power source, so as to simulate the traditional AVC control method. The calculation results are shown in Table 1, and the convergence curve of the fitness value of the objective function is shown in Table 1. Figure 2 As shown.

[0196] Table 1

[0197]

[0198] As shown in Table 1, all schemes control the bus voltage within the upper and lower limits. However, Scheme 1 reduces the active power loss in the station compared to Scheme 2. Therefore, this method can reduce the active power loss of the photovoltaic power station to a certain extent while ensuring that the bus voltage is regulated in place. Compared with the traditional AVC control method, it improves the economic efficiency of the power station operation.

[0199] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A photovoltaic power plant AVC control method based on particle swarm optimization algorithm, characterized in that, Includes the following steps: The voltage U of the high-voltage busbar of the main transformer is set through the automatic voltage control (AVC) system of the photovoltaic power station. pcc_ref ; A multi-objective optimization model is established with the objectives of minimizing the voltage deviation of the high-voltage side busbar of the main transformer in the photovoltaic power station and minimizing the active power loss within the station. Using fuzzy theory, a multi-objective model is transformed into a single-objective model; The particle swarm optimization algorithm is used to optimize and solve the single-objective model; Based on the solution results, the AVC system issues corresponding reactive power adjustment commands to each reactive power source. The specific method for establishing a multi-objective optimization model is as follows: (1) With the goal of minimizing the active power loss within the photovoltaic power station, the objective function is expressed as: （1） （2） （3） In the formula: U pcc U is the voltage of the high-voltage side busbar of the main transformer in the photovoltaic power station. pcc_ref This indicates the setpoint of the high-voltage bus voltage on the main transformer side of the photovoltaic power station; U pcc1 The voltage of the low-voltage busbar of the main transformer of the photovoltaic power station is given by: k = number of photovoltaic power generation units; l = number of collector lines; n = number of SVGs; P loss_T P loss_L P represents the active power loss of the power station's main transformer and the transmission line, respectively; loss_t P loss_l These represent the active power losses on the connecting transformer and the collector line, respectively; R T R t R L R l and R SVG_η Let n represent the resistances of the main transformer, connecting transformer, transmission line, collector line, and the ηth SVG, respectively, where η = 1, 2, ..., n; P and Q represent the active power and reactive power generated by the photovoltaic power station, respectively; P i_j Q i_j U represents the active power and reactive power generated by the j-th photovoltaic power generation unit on the i-th collector line, respectively; i_j Q is the output voltage of the j-th photovoltaic power generation unit on the i-th collector line; SVG_η For the reactive power output of the ηth SVG; P i_m Q i_m Let them represent the active power and reactive power generated by the m-th photovoltaic power generation unit on the i-th collector line, respectively; (2) Taking the minimum voltage deviation of the high-voltage side busbar of the main transformer of the photovoltaic power station as the objective, the objective function is expressed as: （4） In the formula: ΔU pcc_max This indicates the maximum allowable deviation of the voltage on the high-voltage side busbar of the main transformer in a photovoltaic power station; (3) The power flow equation constraints are: （5） In the formula: P α Q α Let g represent the active and reactive power injected into node α, respectively; α∈β, representing all nodes connected to node α; g αβ and b αβ U represents the line conductance and susceptance connecting nodes α and β, respectively; α U β θ represents the voltages at nodes α and β, respectively; αβ This represents the phase angle difference between the voltages at nodes α and β. (4) The control variable constraints are: （6） In the formula: Q i_jmax and Q i_jmin Q represents the maximum and minimum reactive power output of the j-th photovoltaic power generation unit on the i-th collector line, respectively; SVG_ηmax and Q SVG_ηmin These represent the maximum and minimum reactive power output of the ηth SVG, respectively; (5) The state variable constraints are: （7） In the formula: U pcc_max U pcc_min These represent the maximum and minimum allowable values ​​of the high-voltage bus voltage of the main transformer, respectively.

2. The photovoltaic power plant AVC control method based on particle swarm optimization algorithm according to claim 1, characterized in that, The multi-objective optimization model includes an objective function that aims to minimize the active power loss within the photovoltaic power station, an objective function that aims to minimize the voltage deviation of the high-voltage side bus of the main transformer of the photovoltaic power station, and constraints; the constraints include power flow equation constraints, control variable constraints, and state variable constraints.

3. The photovoltaic power plant AVC control method based on particle swarm optimization algorithm according to claim 1, characterized in that, The specific method for transforming a multi-objective model into a single-objective model using fuzzy theory is as follows: A single compromise model is established by scaling the active power loss and the voltage deviation of the high-voltage side bus of the main transformer as follows: （8） In the formula: μ1(F1) and μ2(F2) represent the membership values ​​of active power loss and voltage deviation, respectively; μ1 and μ2 are weighting coefficients, and μ1 + μ2 = 1; both active power loss and voltage deviation are minimum objective functions. According to fuzzy theory, the membership functions for μ1(F1) and μ2(F2) are constructed as follows: （9） （10） In the formula: F 1min F 1max These represent the minimum and maximum active power losses of the power station, respectively; F 2min F 2max These represent the minimum and maximum set voltage deviations, respectively.

4. The photovoltaic power plant AVC control method based on particle swarm optimization algorithm according to claim 3, characterized in that, The specific method for optimizing a single-objective model using the particle swarm optimization algorithm is as follows: (1) Determine the algorithm parameters; including population size N, maximum number of iterations T max Particle dimension D, inertia weight ω, and learning factors c1 and c2; (2) Determine the structure X of the particles and initialize the particle swarm; (12) In the formula: M is the number of photovoltaic inverters, Q M This represents the reactive power output of the Mth photovoltaic inverter; (3) Use the Newton-Raphson method to calculate the power flow. Calculate the fitness value minF of each particle according to formula (8). Set the current position of each particle as the individual extreme value. Then evaluate all particles and select the individual with the smallest fitness value as the population extreme value. (4) Update the velocity and position of all particles according to equation (11); At the (t+1)th iteration, the velocity v of the λ-th particle in d dimensions λd and position x λd The update formula is: （11） In the formula: λ=1.2,...,N, where N is the sample population size; d=1.2,...,D, where D is the dimension of the search space; t is the number of iterations; ω is the inertia weight; r1 and r2 are random numbers between [0,1]; c1 and c2 are learning factors; P λd (t) represents the optimal position experienced by the λ-th particle in the d-th iteration; P gd (t) represents the optimal position experienced by all particles in d dimensions at the t-th iteration; v λd (t) represents the d-dimensional velocity of the λ-th particle at the t-th iteration; x λd (t) represents the d-dimensional position of the λ-th particle at the t-th iteration; (5) Recalculate the fitness values ​​of all particles, evaluate them, and update the individual extreme values ​​and the population extreme values; (6) Determine whether t ≥ T max If yes, output the result; otherwise, return to step (4).

Images