Photovoltaic inverter low voltage ride through capability improvement method based on staged control parameter optimization

Abstract

The invention discloses a photovoltaic inverter low voltage ride through capability improvement method based on staged control parameter optimization, and belongs to the technical field of photovoltaic power generation control optimization, and the method comprises the following steps: S1, dividing a transient process into three stages, considering different dynamic characteristics of unit operation corresponding to each stage, and constructing evaluation indexes of each stage; s2, performing correlation analysis on each control parameter and the evaluation index of the photovoltaic grid-connected inverter by adopting a Spearman correlation coefficient analysis method, and determining the control parameter having strong correlation with the evaluation index of each stage; and S3, according to a correlation analysis result, performing staged optimization on the control parameters of the inverter by using a particle swarm algorithm. By the adoption of the method, physical characteristics of the whole low-voltage ride-through process are considered, a low-voltage ride-through process multi-target evaluation system is established, and the relation between the control parameters and the performance of all stages of the transient state is revealed through correlation analysis.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic power generation control optimization technology, and in particular to a method for improving the low voltage ride-through capability of photovoltaic inverters based on staged control parameter optimization. Background Technology

[0002] With the accelerated global energy structure transformation, photovoltaic (PV) power generation is playing an increasingly important role in the power system due to its clean and renewable characteristics. However, PV inverters connect to the grid via power electronics, and their dynamic characteristics differ significantly from those of traditional synchronous motors. When the grid experiences a short-term voltage sag or transient fault, the inverter's low voltage ride-through (LVRT) capability directly affects whether the generator unit can maintain grid connection and provides reactive power support for rapid grid recovery after a fault. If a large number of inverters simultaneously disconnect from the grid due to voltage sags, it will cause more severe voltage fluctuations and may even lead to large-scale regional blackouts.

[0003] Currently, domestic and international scholars' research on optimizing LVRT capabilities mainly falls into two categories: one is optimizing control methods to maximize the converter's capacity; the other is adding hardware circuits to protect critical components. However, existing methods have the following limitations: control strategy optimization often focuses on a single fault stage, lacking a systematic consideration of the entire process, and the complex structure of advanced control algorithms places high demands on the controller's computing power; while hardware protection schemes can improve fault tolerance, they suffer from increased costs, increased size, and insufficient dynamic recovery capabilities after a fault. Summary of the Invention

[0004] The purpose of this invention is to provide a method for improving the low-voltage ride-through capability of photovoltaic inverters based on phased control parameter optimization, which solves the shortcomings of existing photovoltaic inverter low-voltage ride-through capability optimization methods in terms of systematic consideration, control algorithm complexity, and cost-performance balance of hardware protection schemes.

[0005] To achieve the above objectives, this invention provides a method for improving the low-voltage ride-through capability of photovoltaic inverters based on staged control parameter optimization, comprising the following steps:

[0006] S1. Divide the transient process into three stages, consider the different dynamic characteristics of the unit operation in each stage, and construct evaluation indicators for each stage accordingly.

[0007] S2. The Spearman correlation coefficient analysis method is used to conduct correlation analysis between various control parameters and evaluation indicators of photovoltaic grid-connected inverters, and to determine the control parameters that are strongly correlated with the evaluation indicators at each stage.

[0008] S3. Based on the correlation analysis results, the control parameters of the inverter are optimized in stages using the particle swarm optimization algorithm.

[0009] Preferably, S1 specifically includes:

[0010] S11. Divide the transient process into three stages: pre-fault stage A, fault stage B, and post-fault stage C.

[0011] S12. In stage A, based on the key characteristic quantities steady-state voltage deviation and power factor... Establish evaluation indicators F for Phase A pre ;

[0012] S13. Based on the characteristic of stage B exhibiting a voltage drop to the lowest point and maintaining the drop, select the voltage drop depth ΔV and reactive current support capability ΔI. q Based on the key characteristic quantities in the active power fluctuation amplitude ΔP, establish the overall evaluation index F for stage B. dur ;

[0013] S14 and C stages represent the process of voltage recovery from its lowest point to steady state after fault clearance. The recovery speed T is selected accordingly. rec With transient overshoot V out As a key characteristic quantity, the evaluation index F for stage C is established. post ;

[0014] S15. Based on the evaluation indicators of each stage, establish an overall low-voltage ride-through evaluation index:

[0015] F = w A F pre +w B F dur +w C F post ;

[0016] In the formula, w A w B w C This represents the weighting coefficients, and the sum of the three weighting coefficients is 1.

[0017] Preferably, the evaluation index for stage A in S12 is:

[0018]

[0019] In the formula, V steady This is the positive sequence voltage at the grid connection point; V nom Nominal voltage; ΔV max The maximum allowable voltage deviation is defined as follows: w1 and w2 are weighting coefficients, and the sum of the two weighting coefficients is 1.

[0020] Preferably, the overall evaluation index for stage B established in S13 includes:

[0021] Calculate voltage sag depth mapping:

[0022]

[0023] In the formula, V min is the lowest voltage during the voltage drop; k1 is the adjustment coefficient;

[0024] During voltage dips, the dynamic reactive power support capability is evaluated by the reactive current support ratio, and the dynamic reactive power support capability is expressed as follows:

[0025]

[0026] In the formula, I q reactive current amplitude during the drop; I rated Rated current;

[0027] The fluctuation of active power during a fault is represented by the following mapping:

[0028]

[0029] In the formula, P steady k1 represents steady-state active power; k2 is the scaling factor.

[0030] Based on the three mapping results calculated in Phase B, the overall evaluation index for Phase B is established as follows:

[0031]

[0032] Preferably, the evaluation index for stage C in S14 is:

[0033]

[0034] In the formula, T rec After the fault was cleared, the voltage returned to 90% V. nom Time required; T max The maximum acceptable recovery time; V out To restore the maximum voltage overshoot; w3 and w4 are weighting coefficients, and the sum of the two weighting coefficients is 1.

[0035] Preferably, S2 includes:

[0036] S21. Build a low-voltage ride-through control model for photovoltaic inverters and select reactive power control parameter K, voltage outer loop control parameter and current inner loop control parameter.

[0037] S22. Using Spearman's correlation coefficient analysis, the control parameter sequence and the evaluation index sequence are arranged in ascending order to obtain the corresponding rank sequence. The rank difference between the two sequences is calculated, and the correlation coefficient is obtained through calculation.

[0038] S23. Based on the results of the correlation coefficients, determine the control parameters that are strongly correlated with the evaluation indicators of each stage.

[0039] Preferably, the formula for calculating the correlation coefficient in S22 is as follows:

[0040]

[0041] In the formula, ρ represents the Spearman rank correlation coefficient; d i This represents the difference in rank corresponding to the i-th data point; n represents the total number of data samples.

[0042] Preferably, the result of determining the control parameters that are strongly correlated with the evaluation indicators of each stage in S23 is: in stage A, K is optimized first. p3 and K i3 As parameters for region A; in phase B, K and K are prioritized for optimization. p1 and K i1 As a parameter for region B; and as a priority optimization parameter for K in phase C. p2 and K i2 , as a parameter for region C.

[0043] Preferably, S3 includes:

[0044] S31. When using the particle swarm optimization algorithm for parameter optimization, the algorithm is improved to avoid getting trapped in local optima and to accelerate the convergence speed, ensuring that the power system has the best dynamic response performance, as shown in the following formula:

[0045]

[0046] In the formula, c1 and c2 both represent learning factors, T is the maximum number of iterations, and t is the current number of iterations;

[0047] S32. Set the parameters of area A as optimization variables, keep areas B and C unchanged, take the evaluation index of stage A as the objective function, and set the optimization result as A1.

[0048] S33. Set the parameters of area A to A1, set the parameters of area B to optimization variables, keep area C unchanged, take the evaluation index of stage B as the objective function, and set the optimization result to B1.

[0049] S34. The parameters of area A are set to A1, the parameters of area B are set to B1, the parameters of area C are the optimization variables, the evaluation index of stage C is the objective function, and the optimization result is set to C1.

[0050] S35. Taking the comprehensive evaluation index as the final optimization target, set A1, B1 and C1 as optimization variables, and perform iterative optimization to find the optimal output parameters A2, B2 and C2.

[0051] Therefore, the present invention employs the above-mentioned method for improving the low-voltage ride-through capability of photovoltaic inverters based on staged control parameter optimization, which has the following beneficial effects:

[0052] 1. This invention considers the physical characteristics of the entire low-voltage ride-through process, establishes a multi-objective evaluation system for the low-voltage ride-through process, and reveals the relationship between control parameters and transient performance at each stage through correlation analysis;

[0053] 2. This invention improves the adaptive particle swarm optimization algorithm for learning factors, achieving a synergistic improvement in parameter optimization efficiency and global search capability;

[0054] 3. The phased control parameter optimization method proposed in this invention can effectively improve the low voltage ride-through capability of photovoltaic inverters without modifying the inverter topology. It significantly enhances transient performance while ensuring economic efficiency and has certain theoretical guiding significance for engineering practice.

[0055] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0056] Figure 1 This is a schematic diagram of LVRT interval division according to an embodiment of the present invention;

[0057] Figure 2 This is a schematic diagram of the inverter low-voltage ride-through control model according to an embodiment of the present invention;

[0058] Figure 3 This is a schematic diagram illustrating the correlation coefficients between evaluation indicators and control parameters in an embodiment of the present invention.

[0059] Figure 4 This is a flowchart of step S3 in an embodiment of the present invention;

[0060] Figure 5 This is a comparison diagram of positive sequence voltages according to an embodiment of the present invention;

[0061] Figure 6 This is a dynamic reactive power comparison diagram of an embodiment of the present invention;

[0062] Figure 7 This is a diagram illustrating the optimized convergence process in an embodiment of the present invention. Detailed Implementation

[0063] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0064] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings.

[0065] This invention provides a method for improving the low-voltage ride-through capability of photovoltaic inverters based on staged control parameter optimization, comprising the following steps:

[0066] S1. Divide the transient process into three stages, consider the different dynamic characteristics of the unit operation in each stage, and construct evaluation indicators for each stage accordingly.

[0067] S2. The Spearman correlation coefficient analysis method is used to conduct correlation analysis between various control parameters and evaluation indicators of photovoltaic grid-connected inverters, and to determine the control parameters that are strongly correlated with the evaluation indicators at each stage.

[0068] S3. Based on the correlation analysis results, the control parameters of the inverter are optimized in stages using the particle swarm optimization (PSO) algorithm.

[0069] Furthermore, step S1 specifically includes:

[0070] S11. According to NB / T 31066-2015 "Guidelines for Modeling Electrical Simulation Models of Wind Turbine Units", the transient process is divided into three stages: pre-fault stage A, during-fault stage B, and post-fault stage C. Figure 1 As shown.

[0071] S12. In stage A, based on the key characteristic quantities steady-state voltage deviation and power factor... The evaluation indicators for Phase A are as follows:

[0072]

[0073] In the formula, V steady This is the positive sequence voltage at the grid connection point; V nom Nominal voltage; ΔV maxThe maximum allowable voltage deviation is defined by w1 and w2, which are weighting coefficients. The sum of the two weighting coefficients is 1, and their specific values ​​can be selected according to the actual needs of the project; here, both are 0.5.

[0074] S13. Based on the characteristic of stage B exhibiting a voltage drop to the lowest point and maintaining the drop, select the voltage drop depth ΔV and reactive current support capability ΔI. q Based on the key characteristic quantities in the active power fluctuation amplitude ΔP, a total evaluation index for stage B is established, specifically as follows:

[0075] Calculate voltage sag depth mapping:

[0076]

[0077] In the formula, V min is the lowest voltage during the voltage drop; k1 is the adjustment coefficient, which reflects the rate at which the evaluation index decays with the voltage drop depth. The lower the voltage drop, the more drastically the index decreases. Here, k1 is set to 7.5.

[0078] During voltage dips, the dynamic reactive power support capability is evaluated by the reactive current support ratio, and the dynamic reactive power support capability is expressed as follows:

[0079]

[0080] In the formula, I q reactive current amplitude during the drop; I rated This is the rated current.

[0081] The fluctuation of active power during a fault is represented by the following mapping:

[0082]

[0083] In the formula, P steady is the steady-state active power; k2 is the scaling factor, which determines the sensitivity between active power fluctuations and evaluation indicators, and has a value of 0.8.

[0084] Based on the three mapping results calculated in Phase B, the overall evaluation index for Phase B is established as follows:

[0085]

[0086] By using a multiplicative form, the evaluation result will drop rapidly if any one of them is severely deficient.

[0087] S14 and C stages represent the process of voltage recovery from its lowest point to steady state after fault clearance. The recovery speed T is selected accordingly. rec With transient overshoot V out As a key characteristic quantity, the evaluation index for stage C is established as follows:

[0088]

[0089] In the formula, T rec After the fault was cleared, the voltage returned to 90% V. nom Time required; T max The maximum acceptable recovery time; V out To recover the maximum voltage overshoot; w3 and w4 are weighting coefficients, both with a value of 0.5.

[0090] S15. Based on the evaluation indicators of each stage, establish an overall low-voltage ride-through evaluation index:

[0091] F = w A F pre +w B F dur +w C F post ;

[0092] In the formula, w A w B w C This represents the weighting coefficient. Different weighting coefficients are added together. Different weighting coefficients can be set according to actual needs to focus on the performance at different stages. Here, the values ​​are 0.3, 0.4, and 0.3 respectively.

[0093] Furthermore, step S2 specifically includes:

[0094] S21. Construct a low-voltage ride-through (LVRT) control model for the photovoltaic inverter, as follows: Figure 2 As shown, reactive power control parameter K and a set of voltage outer loop control parameters K are selected. p1 , K i1 and two sets of current inner loop control parameters K p2 , K i2 , K p3 , K i3 .

[0095] S22, such as Figure 3 As shown, using Spearman's correlation coefficient analysis, the control parameter series and the evaluation index series are arranged in ascending order to obtain the corresponding rank series. The rank difference between the two series is calculated, and the correlation coefficient is obtained by substituting it into the following formula:

[0096]

[0097] In the formula, ρ represents the Spearman rank correlation coefficient; d iThis represents the difference in rank corresponding to the i-th data point; n represents the total number of data samples. This invention quantifies the influence of control parameters on indicators by calculating the Spearman correlation coefficient. A coefficient close to +1 indicates a positive correlation, close to -1 indicates a negative correlation, and close to 0 indicates no correlation.

[0098] S23. Based on the correlation coefficients of various control parameters and evaluation indicators, it is analyzed that: in stage A, K should be optimized first. p3 and K i3 That is, the parameters in area A; in stage B, K and K should be optimized first. p1 and K i1 That is, the parameters in region B; in stage C, K should be optimized first. p2 and K i2 That is, the parameters of region C.

[0099] Furthermore, referring to Figure 4 Step S3 is as follows:

[0100] S31. When using the PSO algorithm for parameter optimization, improvements are made to avoid getting trapped in local optima and to accelerate convergence, thereby ensuring that the power system possesses optimal dynamic response performance. As shown in the following equation:

[0101]

[0102] In the formula, c1 and c2 both represent learning factors, T is the maximum number of iterations, and t is the current number of iterations. As the iteration progresses, c1 first increases and then decreases, biased towards the individual optimum, which is beneficial for global search; c2 first decreases and then increases, biased towards the global optimum, which is beneficial for local search and algorithm convergence.

[0103] S32. Set the parameters of area A as optimization variables, while keeping areas B and C unchanged. Use the evaluation index of stage A as the objective function, and set the optimization result as A1.

[0104] S33. Set the parameters of area A to A1, set the parameters of area B to optimization variables, leave area C unchanged, take the evaluation index of stage B as the objective function, and set the optimization result to B1.

[0105] S34. The parameters for area A are set to A1, the parameters for area B are set to B1, the parameters for area C are optimization variables, the evaluation index for stage C is used as the objective function, and the optimization result is set to C1.

[0106] S35. Taking the comprehensive evaluation index as the final optimization target, A1, B1, and C1 are set as optimization variables, and iterative optimization is performed within the range of [0.9(A1B1C1), 1.1(A1B1C1)] to find the optimal output parameters A2, B2, and C2.

[0107] Example

[0108] To verify the effectiveness of the proposed phased parameter optimization method, a photovoltaic grid-connected low-voltage ride-through model was built in MATLAB / Simulink. The short-circuit ratio was set to 1.5, and the grid-side voltage dropped to 0.2 pu for 0.625 s. The control parameters were optimized under this condition.

[0109] Comparison of positive sequence voltage and dynamic reactive power before and after optimization, for example Figure 5 and Figure 6 As shown, in the steady-state range before the fault, the optimized voltage deviation is smaller; during the fault, the voltage drop is improved due to the enhanced dynamic reactive power support capability; in the recovery phase, not only is the recovery time reduced, but the overshoot of voltage recovery is also reduced.

[0110] Table 1 Optimization Results

[0111]

[0112] The results of the indicators before and after optimization are listed in Table 1. The indicators improved in all stages of the low-voltage transient test. In stage B, the operating condition was a grid-side voltage drop to 0.2 pu, which resulted in a low evaluation indicator due to the excessively low voltage drop. After optimization, the dynamic reactive power support capability was enhanced, the voltage drop was improved, and the evaluation indicator increased by 10.08%. The overall evaluation indicator improved by 2.75%, demonstrating that the proposed method can effectively improve the inverter's low-voltage ride-through capability.

[0113] To further verify the phased parameter optimization method and improve the reliability of the PSO algorithm, two algorithms, the traditional Particle Swarm Optimization (PSO) algorithm and the Grey Wolf Optimization (GWO) algorithm, were set as control groups. The initial population size was set to 30 and the number of iterations was 50. Figure 7 The optimization and convergence processes of the three algorithms are shown. For ease of viewing, 1-F is used as the ordinate, with a smaller fitness value indicating a higher overall evaluation score. It can be seen that the improved PSO algorithm has the smallest initial value, converges quickly, and exhibits the strongest local search capability due to the adaptive adjustment of the learning factor, demonstrating significant advantages in both optimization accuracy and convergence speed.

[0114] Table 2 Optimization results of different methods

[0115]

[0116] Table 2 shows the optimization results of different methods. The two traditional algorithms perform overall optimization without targeting individual parameters, resulting in insignificant improvements or even a decline in performance metrics at certain stages. For example, the performance metrics in stage A under the PSO algorithm are lower than before optimization. The IPSO algorithm, however, performs staged optimization, balancing targeted optimization of individual parameters with overall performance improvement. The comprehensive evaluation metrics are improved by 2.34% and 0.97% compared to PSO and GWO, respectively.

[0117] Therefore, the present invention adopts the above-mentioned method for improving the low voltage ride-through capability of photovoltaic inverters based on phased control parameter optimization, which can effectively improve the low voltage ride-through capability of photovoltaic inverters without modifying the inverter topology. It significantly enhances transient performance while ensuring economy and has certain theoretical guiding significance for engineering practice.

[0118] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for improving the low-voltage ride-through capability of photovoltaic inverters based on staged control parameter optimization, characterized in that, Includes the following steps: S1. Divide the transient process into three stages, consider the different dynamic characteristics of the unit operation in each stage, and construct evaluation indicators for each stage accordingly. S2. The Spearman correlation coefficient analysis method is used to conduct correlation analysis between various control parameters and evaluation indicators of photovoltaic grid-connected inverters, and to determine the control parameters that are strongly correlated with the evaluation indicators at each stage. S3. Based on the correlation analysis results, the control parameters of the inverter are optimized in stages using the particle swarm optimization algorithm.

2. The method for improving the low-voltage ride-through capability of a photovoltaic inverter based on staged control parameter optimization according to claim 1, characterized in that, S1 specifically includes: S11. Divide the transient process into three stages: pre-fault stage A, fault stage B, and post-fault stage C. S12. In stage A, based on the key characteristic quantities steady-state voltage deviation and power factor cosφ, establish the stage A evaluation index F. pre ; S13. Based on the characteristic of stage B exhibiting a voltage drop to the lowest point and maintaining the drop, select the voltage drop depth ΔV and reactive current support capability ΔI. q Based on the key characteristic quantities in the active power fluctuation amplitude ΔP, establish the overall evaluation index F for stage B. dur ; S14 and C stages represent the process of voltage recovery from its lowest point to steady state after fault clearance. The recovery speed T is selected accordingly. rec With transient overshoot V out As a key characteristic quantity, the evaluation index F for stage C is established. post ; S15. Based on the evaluation indicators of each stage, establish an overall low-voltage ride-through evaluation index: F=w A F pre +w B F dur +w C F post ; In the formula, w A w B w C This represents the weighting coefficients, and the sum of the three weighting coefficients is 1.

3. The method for improving the low-voltage ride-through capability of a photovoltaic inverter based on staged control parameter optimization according to claim 2, characterized in that, The evaluation indicators for stage A in S12 are: In the formula, V steady This is the positive sequence voltage at the grid connection point; V nom Nominal voltage; ΔV max The maximum allowable voltage deviation is defined as follows: w1 and w2 are weighting coefficients, and the sum of the two weighting coefficients is 1.

4. The method for improving the low-voltage ride-through capability of a photovoltaic inverter based on staged control parameter optimization according to claim 3, characterized in that, The overall evaluation indicators for Phase B established in S13 include: Calculate voltage sag depth mapping: In the formula, V min is the lowest voltage during the voltage drop; k1 is the adjustment coefficient; During voltage dips, the dynamic reactive power support capability is evaluated by the reactive current support ratio, and the dynamic reactive power support capability is expressed as follows: In the formula, I q reactive current amplitude during the drop; I rated Rated current; The fluctuation of active power during a fault is represented by the following mapping: In the formula, P steady k1 represents steady-state active power; k2 is the scaling factor. Based on the three mapping results calculated in Phase B, the overall evaluation index for Phase B is established as follows:

5. The method for improving the low-voltage ride-through capability of a photovoltaic inverter based on staged control parameter optimization according to claim 4, characterized in that, The evaluation indicators for stage C in S14 are: Where, T rec After the fault is cleared, the voltage returns to 90% V. nom Time required; T max The maximum acceptable recovery time; V out To restore the maximum voltage overshoot; w3 and w4 are weighting coefficients, and the sum of the two weighting coefficients is 1.

6. The method for improving the low-voltage ride-through capability of a photovoltaic inverter based on staged control parameter optimization according to claim 1, characterized in that, S2 includes: S21. Build a low-voltage ride-through control model for photovoltaic inverters and select reactive power control parameter K, voltage outer loop control parameter and current inner loop control parameter. S22. Using Spearman's correlation coefficient analysis, the control parameter sequence and the evaluation index sequence are arranged in ascending order to obtain the corresponding rank sequence. The rank difference between the two sequences is calculated, and the correlation coefficient is obtained through calculation. S23. Based on the results of the correlation coefficients, determine the control parameters that are strongly correlated with the evaluation indicators of each stage.

7. The method for improving the low-voltage ride-through capability of a photovoltaic inverter based on staged control parameter optimization according to claim 6, characterized in that, The formula for calculating the correlation coefficient in S22 is as follows: In the formula, ρ represents the Spearman rank correlation coefficient; d i This represents the difference in rank corresponding to the i-th data point; n represents the total number of data samples.

8. The method for improving the low-voltage ride-through capability of a photovoltaic inverter based on staged control parameter optimization according to claim 6, characterized in that, The result of determining the control parameters that are strongly correlated with the evaluation indicators of each stage in S23 is as follows: In stage A, K is optimized first. p3 and K i3 As parameters for region A; in phase B, K and K are prioritized for optimization. p1 and K i1 As a parameter for region B; and as a priority optimization parameter for K in phase C. p2 and K i2 , as a parameter for region C.

9. The method for improving the low-voltage ride-through capability of a photovoltaic inverter based on staged control parameter optimization according to claim 8, characterized in that, S3 includes: S31. When using the particle swarm optimization algorithm for parameter optimization, the algorithm is improved to avoid getting trapped in local optima and to accelerate the convergence speed, ensuring that the power system has the best dynamic response performance, as shown in the following formula: In the formula, c1 and c2 both represent learning factors, T is the maximum number of iterations, and t is the current number of iterations; S32. Set the parameters of area A as optimization variables, keep areas B and C unchanged, take the evaluation index of stage A as the objective function, and set the optimization result as A1. S33. Set the parameters of area A to A1, set the parameters of area B to optimization variables, keep area C unchanged, take the evaluation index of stage B as the objective function, and set the optimization result to B1. S34. The parameters of area A are set to A1, the parameters of area B are set to B1, the parameters of area C are the optimization variables, the evaluation index of stage C is the objective function, and the optimization result is set to C1. S35. Taking the comprehensive evaluation index as the final optimization target, set A1, B1 and C1 as optimization variables, and perform iterative optimization to find the optimal output parameters A2, B2 and C2.

Images