A voltage-current dual-loop adaptive robust control method for parallel inverter systems

Abstract

This invention proposes a dual-loop adaptive robust control method for voltage and current in a parallel inverter system. The method includes the following steps: first, for a multivariable, strongly coupled inverter dynamic model, an adaptive gain-based extended state observer is designed to reduce the initial estimated peak value to estimate the system's lumped disturbance; second, an outer-loop voltage sliding-mode control of the inverter system, based on the above-mentioned observer compensation, is designed to enable the output voltage to track its set reference value with high precision; and finally, an inner-loop current adaptive proportional-integral control law based on a nonlinear function is designed to meet the inner-loop control requirements. This dual-loop adaptive control strategy, combining an outer-loop voltage SMC based on AGESO and an inner-loop current adaptive PI, ensures stable operation of the parallel inverter system, enhances its ability to withstand disturbances such as load and filter parameters, improves the system's dynamic performance, and enables faster equalization of load power among the parallel inverters.

Description

Technical Field

[0001] The present invention belongs to the field of microgrid system control based on new energy, and relates to a voltage-current dual-loop adaptive robust control method for a parallel inverter system. Background Art

[0002] Distributed renewable energy sources are typically connected to microgrids through inverters. However, the capacity of a single inverter is limited, and designing a single large-capacity inverter is difficult and uneconomical. To meet the needs of microgrid expansion, parallel inverters are widely used in practical projects. Therefore, to improve the reliability and power quality of microgrids, studying the control strategy of parallel inverter systems has important theoretical significance and engineering value.

[0003] The control strategy of parallel inverter systems currently generally adopts single-loop voltage or current control. Among them, the single-loop control law is simple, but its dynamic response is slow and it is difficult to cope with various disturbances such as load fluctuations and changes in filter parameters. To overcome this shortcoming and simultaneously achieve power sharing of the parallel inverter system, it is necessary to simultaneously control the outer loop voltage and inner loop current of the system, which is referred to as a dual-loop control strategy.

[0004] For the voltage outer loop in the dual-loop control strategy, a linear proportional integral (PI) controller is mostly used. However, the system dynamic performance of this control method is poor, especially when external disturbances occur, the distortion rate of the system output voltage waveform is high. Therefore, nonlinear control methods such as repetitive control, fuzzy logic control, model-free predictive control, and sliding mode control (SMC) have emerged to overcome the above shortcomings. Due to its strong robustness, SMC has been widely used in voltage outer loop control.

[0005] Conventional SMC strategies only provide robustness in sliding mode, but remain sensitive to disturbances encountered during the arrival phase before entering sliding mode. Therefore, it is necessary to investigate ways to enhance the global robustness of SMC. To this end, this paper first constructs an adaptive observer to estimate the lumped disturbance during system operation in real time, and then proposes an outer-loop voltage SMC strategy based on compensation for this lumped disturbance.

[0006] The Extended State Observer (ESO), the core of active disturbance rejection control (ADRC), estimates system state variables and disturbances in real time and has been widely used in control engineering. Due to the significant deviation between the estimated initial values ​​and the actual values, traditional ESOs suffer from an "initial estimate peak" phenomenon. To address this issue, an adaptive gain ESO (AGESO) can be used. Therefore, this paper proposes an AGESO-based outer-loop voltage SMC control strategy to enhance the system's global robustness and improve its dynamic performance.

[0007] For the current inner loop in the dual-loop control strategy, a linear PI controller is generally used in actual engineering applications. Existing technicians have proposed a current loop control method that combines PI with repetitive control to enhance the dynamic characteristics of the system. However, due to the periodic delay in the repetitive control itself, it leads to large system error fluctuations during actual operation, and has the defects of poor dynamic performance and high sensitivity to parameter disturbances. In addition, existing technicians have proposed a current loop control method that combines PI with model predictive control, and compensates for the disturbances caused by parameter changes by adding a control structure, but its design process is complicated. Therefore, the present invention proposes a voltage and current dual-loop adaptive robust control method for a parallel inverter system to solve the above-mentioned technical defects. Summary of the Invention

[0008] The present invention proposes a voltage-current dual-loop adaptive robust control method for a parallel inverter system, which makes the parallel inverter system more robust and anti-disturbance capable, so as to better adapt to load disturbances and filter inductance parameter perturbations. At the same time, it can also solve the problem of large deviation between the initial value of traditional ESO estimation and the actual value.

[0009] The present invention specifically comprises the following steps:

[0010] Step 1: Based on the dq axis rotating coordinate system, construct the dynamic mathematical model of the inverter circuit. The dynamic mathematical model of the inverter circuit is shown as follows:

[0011] ;

[0012] ;

[0013] Where, For the inverter Inductor current below the axis; For the inverter Inductor current below the axis; is the inverter bridge arm midpoint voltage after coordinate transformation Axis component; is the q-axis component of the inverter bridge arm midpoint voltage after coordinate transformation; For the inverter Output voltage under the shaft; For the inverter Output voltage under the shaft; For the inverter Output current under the shaft; For the inverter Output current under the shaft; For time; is the inductance of the filter; is the capacitance of the filter; is the equivalent resistance of the filter; is the angular frequency of the alternating current in the stationary coordinate system.

[0014] Step 2: Design an SMC control law for the outer loop voltage of the inverter system based on AGESO compensation to enable the output voltage to track its set reference value with high precision; specifically, the following steps are involved:

[0015] Construct an inverter dynamic model based on AGESO compensation. The inverter dynamic model is shown as follows:

[0016] ;

[0017] Where, is the first state variable The derivative of is the second state variable The derivative of the first state variable ; The second state variable ; is the outer loop voltage SMC control law;

[0018] The outer loop voltage SMC control law based on AGESO is designed. The outer loop voltage SMC control law is shown as follows:

[0019]

[0020] Where, is the positive sliding mode coefficient; and are all positive real numbers; is the second state variable Observed values ​​of is a symbolic function; is the sliding surface; is the first state variable Observed values ​​of

[0021] in, The expression is as follows:

[0022] ;

[0023] ;

[0024] Where, for The error between the shaft reference voltage and its observed output voltage; for The derivative of For inverter systems Reference voltage of the axis; AGESO observations Shaft voltage.

[0025] Stability analysis of the outer loop voltage SMC control law based on AGESO:

[0026] Construct the Lyapunov function as shown below:

[0027] ;

[0028] in, Refers to the synovial surface, which is a function of time;

[0029] In order to effectively suppress chattering, the exponential reaching law is selected :

[0030] ;

[0031] Where, and are all positive real numbers;

[0032] For Lyapunov function The derivative is shown below:

[0033] ;

[0034] Obviously, is a positive definite function; since 、 are greater than 0, so It is a negative definite function. Therefore, according to the Lyapunov stability theory, the outer loop voltage SMC control law based on AGESO can ensure the stability of the outer loop voltage system.

[0035] Step 3: Design based on nonlinearity The inner loop current adaptive PI control law of the function is used to improve the system's adaptability to parameter perturbations and speed up the system response;

[0036] Design based on The inner loop current of the function is adaptive PI control law, and the PI control law is shown as follows:

[0037] ;

[0038] Where, It is the inner loop current adaptive PI control law; for The error of the shaft inductance current;

[0039] in, The expression is as follows:

[0040] ;

[0041] Where, For the inverter The reference value of the inductor current under the axis;

[0042] in, The expression is as follows:

[0043] ;

[0044] in, The expression is as follows:

[0045] ;

[0046] Where, is the constant proportional gain; is the piecewise nonlinear function in the proportional term; It is a constant in the proportional term, and its value range is 0-1; is a positive constant in the proportional term; is the constant integral gain; is the piecewise nonlinear function in the integral term; is a constant in the integral term, with a value range of 0-1; is a positive constant in the integral term.

[0047] based on The inner loop current adaptive PI control law of the function can ensure that: when the current error is large, the proportional gain and integral gain in the PI control will become smaller, thereby accelerating the dynamic response of the system; when the current error decreases, the proportional gain and integral gain in the PI control will become larger, thereby reducing the system steady-state error and improving the system steady-state accuracy. Therefore, based on The inner loop current adaptive PI control law of the function makes the system adaptable to load and parameter disturbances.

[0048] The beneficial effects of the present invention are:

[0049] The present invention has the ability to enhance the parallel inverter system's ability to resist load disturbances and filter parameter perturbations, specifically:

[0050] Firstly, for the multivariable and strongly coupled inverter dynamic model, the present invention designs an AGESO that can reduce the initial estimation peak to estimate the system's lumped disturbance;

[0051] Secondly, the present invention designs an inverter system outer loop voltage SMC based on AGESO compensation, so that the output voltage can track its set reference value with high precision;

[0052] Finally, the present invention designs a nonlinear The inner loop current adaptive PI control law of the function is used to meet the inner loop control requirements of fastness, high precision and robustness.

[0053] Specifically:

[0054] 1. The present invention can accelerate the response speed of the system, has a good control effect on the parallel inverter system, has high quality output voltage and current waveforms, and has a fast power sharing speed;

[0055] 2. Compared with the traditional PI dual-loop control, the present invention makes the parallel inverter system more robust and anti-disturbance capable, and can better adapt to load disturbances and filter inductance parameter perturbations;

[0056] 3. The present invention can quickly track the inverter output voltage, has a fast observation error convergence speed, and can reduce the initial estimation peak value, with a faster estimation speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 is a control block diagram of the parallel inverter system in this embodiment;

[0058] Figure 2 This is a block diagram of the outer loop voltage SMC control based on AGESO in this embodiment;

[0059] Figure 3 In this embodiment, it is based on The inner loop current adaptive PI control block diagram of the function;

[0060] Figure 4 It is the three-phase voltage and three-phase current response waveform of the first inverter in system 1;

[0061] Figure 5 This is the three-phase voltage and three-phase current response waveform of the second inverter in system 1;

[0062] Figure 6 This is a schematic diagram comparing the active power of the first inverter and the second inverter in system 1;

[0063] Figure 7 This is a schematic diagram comparing the reactive power of the first inverter and the second inverter in system 1;

[0064] Figure 8 Schematic diagram comparing the d-axis output voltage and d-axis output current of system 1 and system 2 under load disturbance;

[0065] Figure 9 It is a schematic diagram comparing the active power of system 1 and system 2 under load disturbance;

[0066] Figure 10 It is a comparative diagram of the reactive power of system 1 and system 2 under load disturbance;

[0067] Figure 11 This is a schematic diagram comparing the output voltage and output current of phase a of system 1;

[0068] Figure 12 This is a schematic diagram comparing the output voltage and output current of phase a of system 2;

[0069] Figure 13 It is System 1 Estimates and System 3 Schematic diagram of comparison of estimated values;

[0070] Figure 14 It is System 1 Estimation Error and System 3 Schematic diagram of estimation error comparison;

[0071] Figure 15 This is a schematic diagram comparing the active power of system 1 and system 3;

[0072] Figure 16 It is a comparison diagram of reactive power between system 1 and system 3;

[0073] Figure 17 Schematic diagram comparing the d-axis output voltage and d-axis output current of system 1 and system 2 under load disturbance;

[0074] Figure 18 is a schematic diagram of the output voltage and output current of phase a of system 1 when the filter inductance parameters are perturbed;

[0075] Figure 19 Schematic diagram of the output voltage and output current of phase a of system 2 when the filter inductance parameters are perturbed;

[0076] Figure 20 It is System 1 Estimates and System 3 Schematic diagram of comparison of estimated values;

[0077] Figure 21 It is System 1 Estimation error and ESO Schematic diagram of the comparison of estimation errors. DETAILED DESCRIPTION

[0078] The following is a clear and complete description of the invention in conjunction with the figures and implementation methods.

[0079] Example

[0080] like Figure 1 As shown, for a parallel inverter system formed by connecting two identical three-phase voltage source inverters in parallel, in order to overcome load disturbances and filter parameter changes, effectively improve the robustness of the system output voltage and output current control, and achieve rapid equal distribution of load power among the inverters, this embodiment proposes a voltage and current dual-loop adaptive robust control method for the parallel inverter system, including the following steps:

[0081] Step 1: Based on the dq axis rotating coordinate system, construct the dynamic mathematical model of the inverter circuit. The dynamic mathematical model of the inverter circuit is shown as follows:

[0082] ;

[0083] ;

[0084] Where, For the inverter Inductor current below the axis; For the inverter Inductor current below the axis; is the inverter bridge arm midpoint voltage after coordinate transformation Axis component; is the q-axis component of the inverter bridge arm midpoint voltage after coordinate transformation; For the inverter Output voltage under the shaft; For the inverter Output voltage under the shaft; For the inverter Output current under the shaft; For the inverter Output current under the shaft; For time; is the inductance of the filter; is the capacitance of the filter; is the equivalent resistance of the filter; is the angular frequency of the alternating current in the stationary coordinate system.

[0085] Step 2: Figure 2As shown in the figure, an SMC control law for the outer loop voltage of the inverter system based on AGESO compensation is designed to make the output voltage track its set reference value with high precision. Specifically, it includes:

[0086] Construct an inverter dynamic model based on AGESO compensation. The inverter dynamic model is shown as follows:

[0087] ;

[0088] is the first state variable The derivative of is the second state variable The derivative of the first state variable ; The second state variable ; is the outer loop voltage SMC control law;

[0089] The outer loop voltage SMC control law based on AGESO is designed. The outer loop voltage SMC control law is shown as follows:

[0090] ;

[0091] Where, is the positive sliding mode coefficient; and are all positive real numbers; is the second state variable Observed values ​​of is a symbolic function; is the sliding surface; is the first state variable Observed values ​​of

[0092] in, The expression is as follows:

[0093] ;

[0094] ;

[0095] Where, for The error between the shaft reference voltage and its observed output voltage; for The derivative of For inverter systems Reference voltage of the axis; AGESO observations Shaft voltage.

[0096] Step 3: Figure 3 As shown, the design is based on nonlinear The inner loop current adaptive PI control law of the function is used to improve the system's adaptability to parameter perturbations and speed up the system response.

[0097] In step 3, based on The inner loop current of the function is adaptive PI control law, and the PI control law is shown as follows:

[0098] ;

[0099] Where, It is the inner loop current adaptive PI control law; for The error of the shaft inductance current;

[0100] in, The expression is as follows:

[0101] ;

[0102] Where, For the inverter The reference value of the inductor current under the axis;

[0103] in, The expression is as follows:

[0104] ;

[0105] in, The expression is as follows:

[0106] ;

[0107] Where, is the constant proportional gain; is the piecewise nonlinear function in the proportional term; It is a constant in the proportional term, and its value range is 0-1; is a positive constant in the proportional term; is the constant integral gain; is the piecewise nonlinear function in the integral term; is a constant in the integral term, with a value range of 0-1; is a positive constant in the integral term.

[0108] To verify the effectiveness and superiority of this embodiment, three parallel inverter systems were built on the MATLAB / Simulink simulation platform, namely system 1, system 2 and system 3, and experiments 1 to 4 were completed.

[0109] For the outer loop voltage and inner loop current, system 1 adopts the control method provided in this embodiment, system 2 adopts the conventional proportional integral and proportional integral control method, and system 3 adopts sliding mode control based on conventional extended state observer and adaptive proportional integral based on Nfal.

[0110] System 1, System 2 and System 3 all include two inverters with the same capacity and filter parameters, and their line impedances are and , and a droop controller with exactly the same parameters is used. The remaining system parameters are shown in Table 1, and the controller parameters are shown in Table 2.

[0111] Table 1 System parameter settings

[0112]

[0113] Table 2 Controller parameter settings

[0114]

[0115] Figure 1 The load 1 and load 2 in the simulation are both resistive and inductive loads. The active and reactive power of load 1 are 20kW and 20kVar respectively, and the active and reactive power of load 2 are 10kW and 10kV respectively. The simulation time is set to 3s. In the time period of 0s-1s, load 1 is connected to the system. In the time period of 1s-2s, Figure 1 The middle switch S is closed, and load 2 is connected to the system; in the 2s-3s period, Figure 1 The middle switch S is opened, and the load 2 exits the system. Experiments 1 to 7 are all carried out under this working condition.

[0116] Experiment 1: Verify the three-phase voltage and current control effect of system 1:

[0117] like Figure 4 and 5 As shown in the figure, when the load is loaded for 1s or the load is reduced for 2s, the three-phase voltage waveforms of the two inverters do not change significantly, and their amplitudes are stable at 311V. The three-phase output current amplitudes of the two inverters increase rapidly from 30.5A to 45.6A when the load is loaded for 1s, with a response time of about 0.03s. When the load is reduced for 2s, the three-phase output current amplitudes of the two inverters increase rapidly from 30.5A to 45.6A when the load is loaded for 1s, with a response time of about 0.03s. The three-phase output current amplitudes of the two inverters increase rapidly from 30.5A to 45.6A when the load is reduced for 2s, with a response time of about 0.03s. The above data show that system 1 has strong adaptability to load disturbances and enables the parallel inverter system to output higher-quality voltage and current waveforms.

[0118] Depend on Figure 6 and 7As shown in the figure, the active power and reactive power of system 1 reach their steady-state values ​​at 0.15s and 0.25s after startup, respectively. The steady-state active power and reactive power of each inverter are 10kW and 10kVar, respectively. When loading for 1s or unloading for 2s, the active power and reactive power of the two inverters reach exactly the same steady-state values ​​again after about 0.12s. The above data show that the droop control of system 1 can ensure that the load power is evenly distributed between the two inverters.

[0119] Experiment 2: Comparison of control effects between system 1 and system 2 under load disturbance:

[0120] The outer loop voltage and inner loop current of system 1 and system 2 use exactly the same PI control parameters. The proportional gain and integral gain of the inner loop current PI controller of system 2 are the same as those of the inner loop current adaptive controller of system 1. and Exactly the same.

[0121] Depend on Figure 8 As shown in the figure, regardless of loading or unloading, the d-axis output voltage of system 1 and system 2 is basically stable at 311V, while the d-axis output current of system 1 reaches the new steady-state value faster than that of system 2. For example, when the d-axis output current increases from 21.2A to the steady-state value of 31.6A during 1s loading, it takes about 0.02s for system 1 and 0.1s for system 2, and the d-axis output current jitter of system 2 is very obvious. The above data show that system 1 has stronger robustness than system 2.

[0122] Depend on Figure 9 and 10 As shown in the figure, regardless of loading or unloading, the output power of system 1 reaches the new steady-state value faster than that of system 2. For example, when loading for 1 second, the time it takes for the active power of system 1 to reach the new steady-state value of 15,000 W is 0.15 seconds, and the time it takes for the reactive power of system 1 to reach the new steady-state value of 15,000 Var is 0.12 seconds. The time it takes for the active power of system 2 to reach the new steady-state value of 15,000 W is 0.3 seconds, and the time it takes for the reactive power of system 2 to reach the new steady-state value of 15,000 Var is 0.32 seconds. The above data show that system 1 is more capable of resisting load interference, enabling the parallel inverters to achieve faster load power balancing, thereby avoiding circulating current between the two inverters and improving the power quality of the parallel inverter system.

[0123] Experiment 3: Comparison of control effects between system 1 and system 2 when filter inductance parameters are perturbed:

[0124] During the simulation process, the filter inductance parameter is 0.7L in the 0.1s-0.2s time period, L in the 0.2s-0.3s time period, and 1.3L in the 0.3s-0.4s time period.

[0125] Depend on Figure 11 and 12 As shown, when the inductance parameter increases or decreases, the The phase output voltage can almost completely track its reference value, while system 2 has a certain degree of jitter around its reference value. Therefore, system 1 is more resistant to interference caused by changes in inductor parameters.

[0126] Experiment 4: Comparison of initial dynamic responses of system 1 and system 3:

[0127] The parameters of the outer loop voltage SMC and inner loop current adaptive PI controllers of system 1 and system 3 are exactly the same.

[0128] Depend on Figure 13 and 14 As shown, the estimated values ​​of the d-axis output voltage of System 1 and System 3 both track its actual value of 311V in steady state, but their initial estimated peak values ​​are approximately 330V and 360V, respectively. Correspondingly, their initial estimation error peak values ​​are approximately 70V and 100V, and their dynamic response times are approximately 0.017s and 0.04s, respectively. The above data show that compared with System 3, System 1 can not only reduce the initial estimation peak value, but also has a faster estimation speed.

[0129] Depend on Figure 15 and 16 As shown in the figure, it is precisely because of the initial estimation peak of the observer that the system has initial fluctuations in active and reactive power. Compared with system 3, the initial fluctuations in active and reactive power of system 1 are significantly smaller. This shows that system 1 based on AGESO has more satisfactory initial dynamic response characteristics.

[0130] In order to further verify the effectiveness and superiority of this embodiment, three parallel inverter systems identical to the above simulation experiments were built on the StarSim / hardware-in-the-loop simulation experiment platform, namely System 1, System 2 and System 3, and experiments five to seven were completed.

[0131] The experimental setup includes: a real-time simulator based on the NI FPGA board in the NI PXI chassis, StarSim software, a TMS320 F28335PGF DSP, an interface box, a personal computer, and an oscilloscope; the sampling period of the controller is set to 50 , and the rest of the parameters are consistent with the above simulation experiments.

[0132] Experiment 5: Comparison of control effects between system 1 and system 2 under load disturbance:

[0133] Figure 17 The olive green line represents System 1 , the blue line represents the system 2 ; Before and after the load change, the system 1 and system 2 It is always stable at about 1.55div×200V / div=310V;

[0134] Figure 17 The green line represents the system 1 , the purple line represents the system 2 ; Before loading, System 1 and System 2 Both are 1.06div×20A / div=21.2A. After loading, the currents of system 1 and system 2 are Both increase to 1.58div×20A / div=31.6A. When the load is reduced, the The above data show that compared with system 2, system 1 has a higher current rating than system 2, regardless of loading or unloading. The number of fluctuations is small and the time to reach the steady-state value is short, so system 1 has a stronger ability to resist load disturbances.

[0135] Experiment 6: Comparison of the control effects of system 1 and system 2 when the filter inductance parameters are perturbed:

[0136] Set working conditions: filter inductance decreases from L to 0.7L in 0.1s and increases to 1.3L in 0.14s;

[0137] Figure 18 and 19 The purple line in the figure represents the output voltage of phase a of the first inverter in system 1. and output current Waveform, the blue line represents the output voltage of phase a of the first inverter in system 2 and output current waveform;

[0138] Depend on Figure 18 and 19 As shown in the figure, before and after the two step changes in the inductance parameters, the output voltage and output current amplitudes of phase a of system 1 and system 2 are approximately 3.11div×100V / div=311V and 1.51div×20A / div=30.5A respectively. The above data show that although the output voltage and output current waveforms of the two are similar, the output voltage and output current amplitudes of system 1 are approximately 3.11div×100V / div=311V and 1.51div×20A / div=30.5A respectively before and after the two step changes in the filter inductance parameters. and The ripple is significantly smaller than that of system 2 and , so system 1 has a stronger ability to resist the disturbance of filter inductance parameters.

[0139] Experiment 7: Comparison of initial dynamic responses of system 1 and system 3:

[0140] Figure 20 Middle: The olive green line represents System 1 about The blue line represents the estimated value of System 3. The purple line represents the estimated value of The actual value of .

[0141] Depend on Figure 20 As shown in the figure, the estimated values ​​of the d-axis output voltage of system 1 and system 3 both track its actual value of 3.11div×100V / div=311V in steady state. However, the initial estimated peak value of system 1 is significantly smaller than that of system 3. The initial estimated peak values ​​of system 1 and system 3 are approximately 3.22div×100V / div=322V and 3.56div×100V / div=356V, respectively.

[0142] Figure 21 Middle: The blue line shows the system 1 about The purple line represents the estimation error of System 3. The estimation error.

[0143] Depend on Figure 21 As shown in the figure, the initial estimation error peak of system 1 is significantly smaller than that of system 3. The initial estimation peak values ​​of the two are approximately 2.05div×50A / div=102.5V and 1.32div×50A / div=66V, respectively. In addition, system 1 has a faster convergence speed of the estimation error.

[0144] The above data show that compared with System 3, System 1 can not only reduce the initial estimation peak, but also has a faster estimation speed.

[0145] Based on the verification results of Experiments 1 to 7, the following conclusions are drawn:

[0146] (1) This embodiment can accelerate the response speed of the system, has a good control effect on the parallel inverter system, has high quality output voltage and current waveforms, and has a fast power sharing speed;

[0147] (2) Compared with the traditional PI dual-loop control, this embodiment makes the parallel inverter system more robust and anti-disturbance capable, and can better adapt to load disturbances and filter inductance parameter perturbations;

[0148] (3) This embodiment can quickly track the inverter output voltage, the observation error converges quickly, and can reduce the initial estimation peak, thus having a faster estimation speed.

Claims

1. A voltage and current dual-loop adaptive robust control method for a parallel inverter system, characterized in that: The following steps are involved: Step 1: Based on the dq axis rotating coordinate system, construct a dynamic mathematical model of the inverter circuit; Step 2: Design the SMC control law of the inverter system outer loop voltage based on AGESO compensation so that the output voltage can track its set reference value with high precision; Step 2 includes the following steps: Construct an inverter dynamic model based on AGESO compensation. The inverter dynamic model is shown as follows: ； Where, is the first state variable The derivative of is the second state variable The derivative of the first state variable ; The second state variable ; is the outer loop voltage SMC control law; The outer loop voltage SMC control law based on AGESO is designed. The outer loop voltage SMC control law is shown as follows: ； Where, is the positive sliding mode coefficient; and are all positive real numbers; is the second state variable Observed values ​​of is a symbolic function; is the sliding surface; is the first state variable Observed values ​​of in, The expression is as follows: ； ； Where, for The error between the shaft reference voltage and its observed output voltage; for The derivative of For inverter systems Reference voltage of the axis; AGESO observations Shaft voltage; Step 3: Design based on nonlinearity The inner loop current adaptive PI control law of the function is used to improve the system's adaptability to parameter perturbations and speed up the system response; based on The inner loop current adaptive PI control law of the function is shown as follows: ； Where, It is the inner loop current adaptive PI control law; for The error of the shaft inductance current; in, The expression is as follows: ； Where, For the inverter The reference value of the inductor current under the axis; in, The expression is as follows: ； in, The expression is as follows: ； Where, is a constant proportional gain; is the piecewise nonlinear function in the proportional term; It is a constant in the proportional term, and its value range is 0-1; is a positive constant in the proportional term; is the constant integral gain; is the piecewise nonlinear function in the integral term; is a constant in the integral term, with a value range of 0-1; is a positive constant in the integral term.

2. The voltage-current dual-loop adaptive robust control method for a parallel inverter system according to claim 1, characterized in that: The circuit dynamic mathematical model of the inverter in step 1 in the dq axis rotating coordinate system is shown as follows: ； ； Where, For the inverter Inductor current below the axis; For the inverter Inductor current below the axis; is the inverter bridge arm midpoint voltage after coordinate transformation Axis component; is the q-axis component of the inverter bridge arm midpoint voltage after coordinate transformation; For the inverter Output voltage under the shaft; For the inverter Output voltage under the shaft; For the inverter Output current under the shaft; For the inverter Output current under the shaft; For time; is the inductance of the filter; is the capacitance of the filter; is the equivalent resistance of the filter; is the angular frequency of the alternating current in the stationary coordinate system.

Images