A method for identifying control parameters of an energy storage converter under low voltage ride through

Abstract

The application discloses a kind of identification methods of control parameter of energy storage converter under low voltage ride through (LVRT).The technical scheme used in the application is as follows: firstly, according to the mathematical model of energy storage converter under low voltage ride through, the time domain response formula of current is established, and the PI control parameters of converter and the control parameters of LVRT process are identified by using particle swarm optimization algorithm, to solve the multi-parameter identification problem of simultaneous identification of linear parameters and nonlinear parameters;Secondly, considering the difference in sensitivity between different parameters, the control parameters of the fault ride-through process are identified in stages, and the fitness function of the particle swarm optimization algorithm is improved, which improves the identification accuracy.

Description

Technical Field

[0001] This invention belongs to the field of new energy power generation control technology, specifically a method for identifying control parameters of energy storage converters under low voltage ride-through. Background Technology

[0002] Efficient, green, and diversified renewable energy has become the direction of energy development today. However, the instability of renewable energy can affect power quality. Energy storage power stations have advantages such as good dynamic response characteristics, long lifespan, and high reliability. Therefore, introducing energy storage technology can effectively solve the problems brought about by renewable energy grid integration and improve power safety. The response characteristics of an energy storage power station are closely related to its correct control model and control parameters. However, actual energy storage power stations often use commercial converters, and the controller of the energy storage power station is in a black box / grey box state, making it impossible to obtain accurate control parameters and control structure, thus making accurate modeling difficult. However, by adopting an identification method based on input and output data, a simulation model of the energy storage converter can be accurately established.

[0003] Currently, domestic and international research on parameter identification for new energy systems mainly focuses on the identification of control parameters and electrical parameters during steady-state operation. However, the control strategies during steady-state operation differ from those during voltage faults, and the low-voltage ride-through process involves numerous parameter identification terms, which exhibit nonlinear characteristics. Therefore, the identification strategies for steady-state operation cannot be directly applied to the low-voltage ride-through process. J. Tao et al.'s paper ["Modeling Principle and Parameter Identification of Double Fed Wind Turbine for Electromechanical Transient Simulation of Power System," 2020 IEEE 4th Conference on Energy Internet and Energy System Integration, (EI2), 2020, pp. 1433-1438] identifies the control parameters of a double-fed wind turbine under low-voltage ride-through based on the PQ decoupling principle. The identification results show that the active current coefficient is constant, while the reactive current coefficient is related to the fault condition. However, it only considers the current coefficient and does not consider the impact of nonlinear parameters on the system when the voltage drop depth is large.

[0004] Because the low-voltage ride-through process involves many parameters, multi-parameter identification can easily lead to low identification accuracy, making it difficult to accurately identify all parameters. Based on the degree of influence of parameters on the system's dynamic performance, the parameters to be identified are divided into high-sensitivity parameters and low-sensitivity parameters. The paper published by Hou Junxian et al. ["The dynamic simulation model and parameter identification method of DFIG type wind generator for power system electro-mechanic simulation," 2014 IEEE PES Asia-Pacific Power and Energy Engineering Conference, 2014, pp. 1-6] classifies the parameters to be identified in the low-voltage ride-through process into three categories: inherent physical parameters, low-sensitivity PI control parameters, and high-sensitivity low-voltage ride-through control parameters. High-sensitivity parameters are used as the parameters to be identified, while low-sensitivity parameters are set as empirical values. This ignores the impact of the accuracy of low-sensitivity PI control parameters on the fault ride-through process. The paper published by Qin Jishuo et al. [Step-by-step identification of control parameters for fault ride-through of grid-connected inverter for permanent magnet wind turbine based on optimization algorithm [J]. Proceedings of the CSEE, 2021, 41(S01):11.] uses particle swarm optimization algorithm to identify the inner loop PI parameters of current and reactive current support coefficient during the low voltage ride-through process of permanent magnet wind turbine. However, it does not consider the situation where the voltage drop depth is large, causing the inverter to enter the nonlinear region, nor does it consider the identification of the current limit value.

[0005] Therefore, existing identification algorithms cannot meet the identification requirements of low voltage ride-through control parameters of energy storage converters, and a new identification method needs to be proposed. Summary of the Invention

[0006] To address the problem of difficulty in identifying system control parameters under low voltage ride-through faults in energy storage converters, this invention provides a method for identifying control parameters of energy storage converters under low voltage ride-through faults. Based on the trajectory sensitivity of parameters, a particle swarm optimization algorithm for phased identification is proposed to accurately identify the control parameters of energy storage converters during low voltage ride-through.

[0007] The technical solution adopted in this invention is: a method for identifying control parameters of an energy storage converter under low voltage ride-through conditions, the identification steps of which are as follows:

[0008] 1) Based on the control strategy of the low voltage ride-through process of the energy storage converter, the transfer functions of active and reactive currents are obtained, and a frequency domain model of the current is established.

[0009] 2) Perform a Laplace transform on the frequency domain model of the current in step 1) and then derive the corresponding time domain model. Determine the particle swarm algorithm for phased identification based on the linear parameters of the time domain model and the nonlinear parameters involved in the low voltage ride-through process.

[0010] 3) Apply disturbances of varying degrees to the grid voltage, and identify the control parameters of the energy storage converter during the low-voltage ride-through process based on the active and reactive current response data. The specific process is as follows:

[0011] 3-a) Perform trajectory sensitivity analysis on the parameters to be identified to determine the sensitivity of each parameter during voltage faults and fault recovery.

[0012] 3-b) Perform the first stage of indiscriminate identification based on the response data of active and reactive currents, and use the identification results as the initial values ​​for the second stage of identification.

[0013] 3-c) Perform a second-stage identification for each control parameter according to the order of trajectory sensitivity from high to low. During the identification, select the sampling data of the response curve segment with the highest trajectory sensitivity. Other parameters are taken as fixed values ​​during the identification process.

[0014] Furthermore, in step 1), the control strategy for the low-voltage ride-through process of the energy storage converter is as follows: when the system detects that the grid voltage is less than 0.9 pu, the power outer loop of the vector control of the energy storage converter is locked, and the current reference value of the current inner loop is no longer given by the output value of the power outer loop, but is directly calculated according to the grid connection regulations; in the fault, the control method of the energy storage converter is reactive current priority control, that is, reactive power is output according to the grid connection regulations, while the active power output is constrained by the current limit and the level of reactive power.

[0015] Furthermore, in step 1), during a voltage fault, the reference values ​​for active and reactive currents are given according to the following formula:

[0016]

[0017]

[0018] Where: k pq2 k iq2 These represent the proportional gain and integral gain of the power control outer loop on the q-axis, respectively, k qv Q is the reactive current support factor. * Here is the reactive power reference value, Q is the reactive power, and i N For the rated current, i max i is the maximum allowable current value of the converter. d0 * i is the reference value for the active current component before the fault. q* i is the reference current along the q-axis. d * is the reference current along the d-axis, s is the Laplace transform operator, and U is the grid voltage.

[0019] Further, in step 1), the frequency domain response of the current is calculated according to the following frequency domain model:

[0020]

[0021]

[0022] Where: k pd1 k id1 These represent the proportional gain and integral gain of the d-axis current inner loop, respectively, and k pq1 k iq1 These represent the proportional gain and integral gain of the q-axis current inner loop, respectively. g This is a filter inductor.

[0023] Further, in step 2), the time-domain response of the current is calculated according to the following time-domain model:

[0024]

[0025]

[0026] Where: k pd1 k id1 These represent the proportional gain and integral gain of the d-axis current inner loop, respectively, and k pq1 k iq1 These represent the proportional gain and integral gain of the q-axis current inner loop, respectively. g This is a filter inductor.

[0027] Furthermore, in step 2), the reference currents of the d and q axes during fault recovery are obtained according to the following formula:

[0028]

[0029]

[0030] Where: k dip k is the active current recovery limit value. diq The reactive current recovery limit value, These are the reference values ​​for the d-axis and q-axis current components during a fault, respectively, and t is the fault recovery time.

[0031] Based on the time-domain model of the d-axis and q-axis currents during fault recovery in step 2), the linear parameters to be identified are: the proportional gain and integral gain k of the d-axis current inner loop.pd1 k id1 The proportional gain and integral gain k of the q-axis current inner loop pq1 k iq1 reactive current support coefficient k qv The nonlinear parameter to be identified is: the maximum allowable current value i of the converter. max Active current recovery limit value k dip Reactive current recovery limit value k diq .

[0032] Furthermore, the energy storage converter adopts a power-current dual closed-loop vector control in steady state, and adopts a strategy of power outer loop blocking, direct current inner loop reference value setting, and reactive current priority control in case of transient faults.

[0033] Furthermore, the particle swarm optimization algorithm transforms the parameter identification problem into a parameter combinatorial optimization problem. Through an efficient search of the entire parameter space, it obtains the optimal parameters that minimize the fitness function, which is:

[0034]

[0035] Where: y r (i) represents the measured active or reactive current data, and y(i) represents the output data of the particle swarm algorithm based on parameter identification.

[0036] To improve the accuracy of parameter identification, the fitness function of the particle swarm optimization algorithm is improved. The goal of parameter optimization is not only to minimize the difference between the actual data and the identified data, but also to minimize the rate of change of both. The improved fitness function is as follows:

[0037]

[0038] Wherein: T s y'(i) represents the sampling time of the measured data, and y'(i) represents the derivative of the output data based on parameter identification.

[0039] Furthermore, in step 3-a), for a linear system, the trajectory sensitivity is calculated by taking the partial derivative of the transfer function:

[0040]

[0041] Where: k dip k is the active current recovery limit value. diq i is the reactive current recovery limit value. max k is the maximum allowable current value of the converter. qv k is the reactive current support factor. pq1 The proportional gain of the inner loop of the q-axis current, k iq1Let k be the integral gain of the inner loop of the q-axis current. pd1 The proportional gain of the inner loop of the d-axis current, k id1 t represents the integral gain of the inner loop of the d-axis current, and t represents the fault recovery time.

[0042] Furthermore, in step 3-a), for the case of a nonlinear system, the trajectory sensitivity is the ratio of the change in the trajectory of the simulated output current response curve to the change in each control parameter when each parameter undergoes a small variation:

[0043]

[0044] Where: P is the output active power, Δk dip This is the disturbance amount for the active current recovery limit value.

[0045] This invention solves the multi-parameter identification problem of simultaneous identification of linear and nonlinear parameters using the particle swarm optimization algorithm; this invention uses trajectory sensitivity analysis to segment the parameters to be identified, achieving accurate identification of all parameters; this invention improves the fitness function of the particle swarm optimization algorithm, thereby improving the identification precision and accuracy.

[0046] The identification method of the present invention is also effective for other new energy units or other control methods, and is especially suitable for multi-parameter identification. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the vector control system for an energy storage converter.

[0048] Figure 2 This is a structural diagram of the low-voltage ride-through model of the energy storage converter identified in this invention;

[0049] Among them, i d i q i represents the d-axis and q-axis components of the current. d_select i q_select These are the selection values ​​for the outer loop outputs of active power and reactive power, respectively; i d * i q * The reference currents for the d and q axes; i d0 * i q0 * These are the reference values ​​for the d-axis and q-axis current components before the fault; i N This is the rated current; i max P represents the maximum allowable current value of the converter; U represents the grid voltage; P and Q represent active and reactive power, respectively; P * Q * For active and reactive power reference values; k pd1 kid1 These represent the proportional gain and integral gain of the d-axis current inner loop, respectively; k pd2 k id2 These represent the proportional gain and integral gain of the d-axis power outer loop, respectively; k pq1 k iq1 These represent the proportional gain and integral gain of the q-axis current inner loop, respectively; k pq2 k iq2 These represent the proportional gain and integral gain of the q-axis power outer loop, respectively; k qv is the reactive current support coefficient; s is the Laplace transform operator; L g This is the filter inductor. Before the fault, the energy storage converter was controlled by a power-current dual closed-loop vector control, with the output of dashed box ① selected as i. d_select and i q_select During the fault, the outer power loop of the energy storage converter is locked, and the outputs of dashed boxes ② and ③ are selected as i respectively. d_select and i q_select Once the fault is cleared and grid voltage recovery is detected, the control mode of the energy storage converter switches back to the pre-fault power-current dual closed-loop control, using the outputs of dashed boxes ④ and ⑤ as i. d_select and i q_select .

[0050] Figure 3 This is a structural diagram of the frequency domain model of the energy storage converter vector control identified in this invention.

[0051] Figure 4 This is an averaged curve of the active power response trajectory sensitivity of the fault ride-through control parameters of this invention.

[0052] Figure 5 This is an averaged curve of the reactive power response trajectory sensitivity of the fault ride-through control parameters of this invention.

[0053] Figure 6 This is a comparison chart of the active power response curves corresponding to the identified value and the actual value when the voltage drops to 0.22 pu under the operating condition of the present invention.

[0054] Figure 7 This is a comparison chart of the reactive power response curves corresponding to the identified value and the actual value when the voltage drops to 0.22 pu under the operating condition of the present invention.

[0055] Figure 8 This is a voltage curve of the present invention when the voltage drops to 0.22 pu.

[0056] Figure 9 This is a comparison chart of the active power response curves corresponding to the identified value and the actual value when the voltage drops to 0.47 pu under the operating condition of the present invention.

[0057] Figure 10This is a comparison chart of the reactive power response curves corresponding to the identified value and the actual value when the voltage drops to 0.47 pu under the operating condition of the present invention.

[0058] Figure 11 This is a voltage curve of the present invention when the voltage drops to 0.47 pu.

[0059] Figure 12 This is a comparison chart of the active power response curves corresponding to the identified value and the actual value when the voltage drops to 0.73 pu under the operating condition of the present invention.

[0060] Figure 13 This is a comparison chart of the reactive power response curves corresponding to the identified value and the actual value when the voltage drops to 0.73 pu under the operating condition of the present invention.

[0061] Figure 14 This is a voltage curve of the present invention when the voltage drops to 0.73 pu. Detailed Implementation

[0062] To describe the present invention in more detail, the technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0063] like Figure 1 The diagram shows the structure of a vector control system for an energy storage converter. The energy storage converter employs grid voltage-oriented vector control, enabling decoupled control of the d-axis and q-axis, thus regulating active and reactive power outputs respectively. This invention identifies the control parameters of the energy storage converter during the low-voltage ride-through process, including the following steps:

[0064] (1) Obtain the transfer functions of active current and reactive current based on the energy storage converter control strategy.

[0065] The control strategy for energy storage converters during steady-state operation differs from that during voltage faults. The fault ride-through process of energy storage converters encompasses three stages: "pre-fault – during fault – fault recovery." These three stages do not entirely employ vector control; rather, they involve switching between control modes within the framework of vector control. The control strategy for energy storage converters during low-voltage ride-through is as follows: Figure 2 As shown. Before the fault, the energy storage converter was controlled by a power-current dual closed-loop vector control system, and the selected... Figure 2 The output of dashed box ① is used as i d_select and i q_select Upon detecting grid voltage fluctuations, the system enters a fault-prone phase. To improve the response speed of the energy storage converter during low-voltage ride-through, the power outer loop of the vector control system is typically locked. The current reference value of the inner current loop is no longer given by the output value of the power outer loop, but is directly calculated according to the grid connection specifications. The outer loop output is determined based on the degree of voltage drop; if the voltage drops above 0.9 pu, then... Figure 2 The output of dashed box ① is used as i d_select and i q_select If the voltage drops to the range of 0.2-0.9 pu, then the outputs of dashed boxes ② and ③ are selected as i respectively. d_select and i q_select In engineering practice, the reference value for the d-axis current component is generally given as a relatively small constant. The basic control method for the energy storage converter during a fault is reactive current priority control, prioritizing reactive power output according to grid connection regulations. Active power output is constrained by current limiting and reactive power level. During a fault, the reference values ​​for active and reactive currents are given as follows:

[0066]

[0067]

[0068] Where: k pq2 k iq2 For the PI parameters of the outer loop of the power control on the q-axis, k qv Q is the reactive current support factor. * Reactive power reference value, i N For the rated current, i max i is the maximum allowable current value of the converter. d0 * This is a reference value for the active current component before the fault.

[0069] Once the fault is cleared and grid voltage recovery is detected, the energy storage converter resumes active power priority control. The reference values ​​for both the d-axis and q-axis current components gradually recover to their pre-fault reference values ​​or to the new post-fault reference values. To avoid the impact of sudden current changes, the recovery speed of the current components is limited by the slope limiting parameter, which switches to... Figure 2 The outputs of dashed boxes ④ and ⑤ are used as i d_select and i q_select The control mode of the energy storage converter is simultaneously switched back to the pre-fault power-current dual closed-loop control, with the active and reactive current reference values ​​set as follows:

[0070]

[0071] Where: k dip and k diq To restore the limiting values ​​for active and reactive currents, This is the reference value for the d-axis current component during a fault. i is the reference value for the q-axis current component during a fault. q0 * t is the reference value of the reactive current component before the fault, and t is the fault recovery time.

[0072] During fault ride-through, only the outer loop control is switched, but the inner current loop remains unchanged, constituting a single closed-loop PI control structure. According to... Figure 3 The control model for the inner current loop can be obtained:

[0073]

[0074]

[0075] Among them: U d U q Let k be the d-axis and q-axis components of the voltage. pd1 k id1 k represents the PI parameters of the inner current loop on the d-axis. pq1 k iq1 These are the PI parameters for the inner current loop on the q-axis.

[0076] Furthermore, the frequency domain model of the inner current loop can be obtained:

[0077]

[0078]

[0079] Where: L g This is a filter inductor.

[0080] (2) Perform a Laplace transform on the frequency domain model of the current to derive the corresponding time domain model:

[0081]

[0082]

[0083] In the event of a grid fault, the outer-loop PI regulator of the energy storage converter's vector control does not require identification due to blocking, but the inner-loop PI parameters of the current loop do. Furthermore, considering that voltage fluctuations at the grid connection point may cause the converter to enter the nonlinear operating region, the relevant control parameters of the limiting circuit also need to be identified. Therefore, the parameters identified using the particle swarm optimization algorithm are: the PI parameters k of the active and reactive power inner-loop current loops. pd1 k id1 k pq1 k iq1 Active and reactive current recovery limiting parameters k dip k diq reactive current support coefficient k qv and current limiting parameter i max .

[0084] The Particle Swarm Optimization (PSO) algorithm transforms the parameter identification problem into a combinatorial optimization problem. Through an efficient search of the entire parameter space, it obtains the optimal parameters that minimize the fitness function, which is:

[0085]

[0086] Where: y r (i) represents the measured active or reactive current data, and y(i) represents the output data of the particle swarm algorithm based on parameter identification.

[0087] During low-voltage ride-through, the system's power response curve undergoes abrupt changes. Traditional particle swarm optimization (PSO) algorithms only minimize the difference between actual and identified data using their fitness function, leading to errors in the identification of control parameters. To improve parameter identification accuracy, this invention modifies the PSO algorithm's fitness function. The goal of parameter optimization is not only to minimize the difference between actual and identified data but also to minimize the rate of change of both. The improved fitness function is as follows:

[0088]

[0089] Wherein: T s y'(i) represents the sampling time of the measured data, and y'(i) represents the derivative of the output data based on parameter identification.

[0090] (3) Different degrees of disturbance were applied to the grid voltage on the experimental platform, and the control parameters of the energy storage converter during the low voltage ride-through process were identified based on the active and reactive current response data. The voltage disturbance amount and disturbance time are shown in Table 1.

[0091] Table 1 Fault Trip Experiment Test Conditions

[0092]

[0093] (3-a) Trajectory Sensitivity Analysis

[0094] To ensure accurate identification of all parameters, this invention performs trajectory sensitivity analysis on the power response during fault ride-through of the energy storage converter, and then identifies each parameter in segments. For linear systems, trajectory sensitivity is calculated by taking the partial derivative of the transfer function:

[0095]

[0096] For nonlinear systems, trajectory sensitivity is the ratio of the change in the trajectory of the simulated output current response curve to the change in each control parameter when the parameters undergo small variations:

[0097]

[0098] Trajectory sensitivity analysis was performed on the parameters to be identified. The trajectory sensitivity curves for each parameter are shown below when the voltage drops between 7 and 9 seconds. Figures 4-5As shown, different parameters have varying degrees of sensitivity at different stages of a fault.

[0099] according to Figures 4-5 It can be seen that the active current response characteristics are affected by the current loop PI parameters and the current limiting parameter i. max Active current recovery slope limiting parameter k dip The combined effects of i max and k dip The reactive current response characteristics are most significantly affected by the current loop PI parameters and the reactive current support coefficient k. qv Current limiting parameter i max Reactive current recovery slope limiting parameter k diq The combined effects of i max and k diq The greatest impact, k qv Secondly, the sensitivity of a parameter changes over time; a highly sensitive parameter only maintains high sensitivity within a certain period. For example, k qv It has high sensitivity during the fault phase, k dip and k diq It has high sensitivity during the fault recovery phase. max It exhibits high sensitivity during both the fault phase and the fault recovery phase.

[0100] (3-b) First-stage parameter identification

[0101] The first stage of identification is based on the sampled data of the current response throughout the fault process. The particle swarm optimization algorithm is used to identify all control parameters involved in the fault ride-through without discrimination, thereby achieving a wide range of optimization of the control parameters and obtaining a set of preliminary identification results.

[0102] (3-c) Second stage parameter identification

[0103] After obtaining a set of initial identification values ​​through the first stage of undifferentiated identification, k dip and k diq First, the power response curve of the fault recovery phase is used to identify the parameters for the second phase. max Then, the power response curves of the fault stage and the fault recovery stage are used to identify the parameters for the second stage, followed by k qv The power response curve of the fault stage is used for the second stage parameter identification. Finally, the current loop PI parameters are used for the second stage parameter identification using the power response curve of the entire stage to obtain the final identification result.

[0104] Based on the measured data such as voltage, active current, and reactive current collected at the grid connection point, the parameters of fault ride-through control were identified, and the final identification results are shown in Table 2.

[0105] Table 2. Fault ride-through control parameter identification results based on experimental data.

[0106]

[0107]

[0108] To verify the accuracy of the identification results, a simulation model of the low-voltage ride-through control system was built on a simulation platform. The control parameters in the simulation model were set to the identification results shown in Table 2. The grid-connected voltage curve, active power response curve, and reactive power response curve obtained from experimental tests and simulation tests based on the identification results are collectively represented in Table 2. Figures 6-14 In the results, the fitting degree of all three sets of curves is good, and the simulation models corresponding to the identified values ​​can well reflect the power response characteristics of the actual converter. The overall power curves under different operating conditions have good fitting degree, and the simulation identification models and actual tests are quite accurate in fitting the step response performance indicators such as rise time, peak time, overshoot, and steady-state value, which confirms the effectiveness of the method of the present invention.

Claims

1. A method for identifying control parameters of an energy storage converter under low voltage ride-through, characterized in that, The identification steps are as follows: 1) Based on the control strategy of the low voltage ride-through process of the energy storage converter, the transfer functions of active and reactive currents are obtained, and a frequency domain model of the current is established. 2) Perform a Laplace transform on the frequency domain model of the current in step 1) to derive the corresponding time domain model. Determine the particle swarm algorithm for phased identification based on the linear parameters of the time domain model and the nonlinear parameters involved in the low voltage ride-through process. 3) Apply disturbances of varying degrees to the grid voltage, and identify the control parameters for the low-voltage ride-through process of the energy storage converter based on the active and reactive current response data. The specific process is as follows: 3-a) Perform trajectory sensitivity analysis on the parameters to be identified to determine the sensitivity of each parameter during voltage faults and fault recovery. 3-b) Perform the first stage of indiscriminate identification based on the response data of active and reactive currents, and use the identification results as the initial values ​​for the second stage of identification. 3-c) Perform the second stage identification for each control parameter according to the order of trajectory sensitivity from high to low. During the identification, select the sampling data of the response curve segment with the highest trajectory sensitivity. Take fixed values ​​for other parameters during the identification process. The Particle Swarm Optimization (PSO) algorithm transforms the parameter identification problem into a combinatorial optimization problem. Through an efficient search of the entire parameter space, it obtains the optimal parameters that minimize the fitness function, which is: in: y r ( i The active or reactive current data is the actual measured value. y ( i This refers to the output data of the particle swarm algorithm based on parameter identification. To improve the accuracy of parameter identification, the fitness function of the particle swarm optimization algorithm is improved. The improved fitness function is as follows: in: T s The sampling time for the measured data. The derivative of the output data based on parameter identification; In step 3-a), for a linear system, the trajectory sensitivity is calculated by taking the partial derivative of the transfer function: in: k dip The active current recovery limit value, k diq For reactive current recovery limiting value, i max This is the maximum allowable current value for the converter. k qv This is the reactive current support factor. k pq1 for q The proportional gain of the inner loop of the shaft current. k iq1 for q Integral gain of the inner loop of the shaft current. k pd1 for d The proportional gain of the inner loop of the shaft current. k id1 for d Integral gain of the inner loop of the shaft current. t This refers to the fault recovery time. In step 3-a), for the case of a nonlinear system, the trajectory sensitivity is the ratio of the change in the trajectory of the simulated output active power response curve to the change in each control parameter when each parameter undergoes a small variation: in: P To output active power, k dip This is the disturbance amount for the active current recovery limit value.

2. The method for identifying control parameters of an energy storage converter under low voltage ride-through according to claim 1, characterized in that, In step 1), the control strategy for the low-voltage ride-through process of the energy storage converter is as follows: When the system detects that the grid voltage is less than 0.9 pu, the power outer loop of the vector control of the energy storage converter is locked, and the current reference value of the current inner loop is no longer given by the output value of the power outer loop, but is directly calculated according to the grid connection regulations; In the fault, the control method of the energy storage converter is reactive current priority control, that is, reactive power is output according to the grid connection regulations, while the active power output is constrained by the current limit and the level of reactive power.

3. The method for identifying control parameters of an energy storage converter under low voltage ride-through according to claim 2, characterized in that, In step 1), during a voltage fault, the reference values ​​for active and reactive currents are given according to the following formula: in: k pq2 , k iq2 They are respectively q The proportional gain and integral gain of the outer loop of the shaft power control. k qv This is the reactive current support factor. Q * This is a reference value for reactive power. Q Reactive power i N Rated current, i max This is the maximum allowable current value for the converter. i d0 * This is a reference value for the active current component before the fault. i q * for q The reference current of the shaft, i d * for d The reference current of the shaft, s For the Laplace transform operator, U This is the grid voltage.

4. The method for identifying control parameters of an energy storage converter under low voltage ride-through according to claim 1, characterized in that, In step 1), the frequency domain response of the current is calculated according to the following frequency domain model: in: k pd1 , k id1 They are respectively d The proportional gain and integral gain of the current inner loop of the shaft. k pq1 , k iq1 They are respectively q The proportional gain and integral gain of the current inner loop of the shaft. L g For filtering inductors, i q * for q The reference current of the shaft, i d * for d The reference current of the shaft, s This is a Laplace transform operator.

5. The method for identifying control parameters of an energy storage converter under low voltage ride-through according to claim 1, characterized in that, In step 2), the time-domain response of the current is calculated according to the following time-domain model: in: k pd1 , k id1 They are respectively d The proportional gain and integral gain of the current inner loop of the shaft. k pq1 , k iq1 They are respectively q The proportional gain and integral gain of the current inner loop of the shaft. L g For filtering inductors, i q * for q The reference current of the shaft, i d * for d The reference current of the shaft, t This refers to the fault recovery time.

6. The method for identifying control parameters of an energy storage converter under low voltage ride-through according to claim 1, characterized in that, In step 2), during the fault recovery period d, q The reference current of the shaft is obtained according to the following formula: in: k dip The active current recovery limit value, k diq For reactive current recovery limiting value, i d * _LVRT , i q * _LVRT When the fault occurs d , q Reference value for shaft current component. t This refers to the fault recovery time. i d0 * , i q0 * Before the fault d , q Reference value of shaft current component Combined with the fault recovery period in step 2) d, q The time-domain model of the shaft current, the linear parameters to be identified are: d The proportional gain and integral gain of the current inner loop of the shaft. k pd1 , k id1 , q The proportional gain and integral gain of the current inner loop of the shaft. k pq1 , k iq1 reactive current support coefficient k qv The nonlinear parameter to be identified is: the maximum allowable current value of the converter. i max Active current recovery limit value k dip Reactive current recovery limit value k diq .

7. The method for identifying control parameters of an energy storage converter under low voltage ride-through according to claim 1, characterized in that, The energy storage converter employs power-current dual closed-loop vector control in steady state, and adopts a strategy of power outer loop blocking, direct current inner loop reference value setting, and reactive current priority control in case of transient faults.

Images