Improved sliding mode variable structure control based off-grid control method for T-type energy storage converter

Abstract

The application discloses a T-type energy storage converter off-grid control method based on improved sliding mode self-disturbance rejection, first obtains a mathematical model in a two-phase rotating coordinate system from a main topological structure of a T-type three-level energy storage converter, and designs a current PI inner ring and a droop outer ring. Then, a mathematical model corresponding to a voltage ring is converted into a first-order self-disturbance rejection paradigm, a linear extended observer with a disturbance differential term as a new state quantity is designed, and is adaptively weighted and connected in parallel with a traditional observer, so that noise at a steady state is reduced, and anti-disturbance capability at a transient state is reserved. Then, a state error feedback control law is designed by using an adaptive improved super-spiral sliding mode control law, so that anti-disturbance capability of the self-disturbance rejection controller is further improved, and inherent chattering is weakened. Finally, the power voltage current three-ring control is output to an SVPWM to control the T-type energy storage converter. Simulation results show that, compared with a traditional method, the method has better dynamic response performance and anti-disturbance performance.

Description

Technical Field

[0001] This invention relates to an off-grid control method for T-type energy storage converters based on improved sliding mode self-disturbance rejection, and belongs to the technical field of off-grid control methods for T-type energy storage converters. Background Technology

[0002] Power conversion systems (PCS) offer advantages such as high-efficiency energy conversion, bidirectional energy flow, and enhanced grid stability. They play a crucial role in solving new energy generation and building microgrids, assisting in smoothing the output of renewable energy, achieving peak shaving and valley filling, and enabling intelligent control. Among them, the T-type three-level converter, with its lower harmonics and lower switching losses, is gradually becoming a key component in new energy generation, distributed energy systems, and microgrids.

[0003] In off-grid operation, droop control is widely used because it is an equivalent control and has the characteristics of plug and play. However, the inner loop of droop control often adopts voltage and current PI dual closed-loop control technology. When faced with the unpredictable switching of microgrid loads, it will cause a large output voltage drop, and its transient process is relatively slow when the power voltage is given, which is not conducive to the frequently changing system environment.

[0004] Active disturbance rejection control (ADRC) has advantages such as strong robustness and is widely used in industrial disturbance suppression. However, traditional ADRC, especially when facing fast and high-frequency disturbances, has difficulty balancing the observation speed and observation noise, and its own observation accuracy is hard to guarantee. Moreover, traditional linear control laws are difficult to achieve optimal control. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the defects of the prior art and provide an off-grid control method for T-type energy storage converters based on improved sliding mode active disturbance rejection, which improves the traditional active disturbance rejection controller to enhance the overall disturbance rejection capability and robustness of the energy storage converter control.

[0006] Preferably, the present invention provides an off-grid control method for a T-type energy storage converter based on improved sliding mode active disturbance rejection, comprising:

[0007] Step 1: Using the main topology of the T-type three-level converter, obtain the mathematical model in a two-phase rotating coordinate system; based on the typical type I system, obtain the proportional coefficient and integral coefficient of the current inner loop, and obtain the power outer loop according to the droop formula;

[0008] Step 2: Transform the formula corresponding to the voltage loop in the mathematical model into the first-order active disturbance rejection paradigm to obtain the second-order linear extended state observer.

[0009] Step 3: Add a perturbation differential term as a new state variable to the second-order linear extended observer obtained in Step 2, construct a new extended-order LESO with perturbation differential term, and add a hysteresis factor at the total perturbation output for initial noise suppression.

[0010] Step 4: Connect the second-order linear extended state observer obtained in Step 2 with the increased-order LESO obtained in Step 3 in a weighted parallel to obtain an adaptively weighted parallel LESO.

[0011] Step 5: An improved adaptive superspiral sliding mode control law is introduced into the state error feedback control law. The hyperbolic tangent function is used to weaken the inherent chattering of the sliding mode, and a voltage loop is constructed based on the information of the lumped disturbance observations provided by the adaptive weighted parallel LESO.

[0012] Step 6: Based on the power outer loop, voltage loop, and current inner loop, obtain the output modulation signal; transmit the output modulation signal to the SVPWM module with center point balancing to control the main topology of the T-type three-level converter.

[0013] Prior to, step 1 includes:

[0014] Using the main topology of a T-type three-level converter, a mathematical model in a two-phase rotating coordinate system is obtained:

[0015]

[0016] In the formula, C f For the AC side LC filter, L is the filter capacitor. f For the filter inductance of the LC filter, i 1d and i 1q These represent the d-axis and q-axis components of the filter inductor current in the main topology filter of a T-type three-level converter, respectively; 2d and i 2q These are the d-axis and q-axis components of the load current on the load side, respectively; u cd and u cq These are the d-axis and q-axis components of the output voltage, respectively. d For the main topology of a T-type three-level converter, the output d-axis voltage, u q The output q-axis voltage of the main topology of the T-type three-level converter is given by R1, which is the parasitic resistance of the filter inductor of the AC side LC filter, and ω is the known angular frequency corresponding to the power frequency. The formula corresponding to the voltage loop is the third and fourth formulas from top to bottom in formula (1).

[0017] Based on the requirement that the voltage angular frequency variation of the power grid should be less than or equal to 1% and the amplitude variation should be less than or equal to 5%, the droop factor of the power outer loop is determined as follows:

[0018]

[0019] In the formula, P max Let ω0 be the maximum active power output of the T-type three-level converter main topology when the frequency decreases, P0 be the rated active power output of the T-type three-level converter main topology, E0 be the rated voltage output of the T-type three-level converter main topology, and Q0 be the rated reactive power output of the T-type three-level converter main topology. min Q is the minimum permissible angular frequency (Q) for the main topology of a T-type three-level converter to output maximum active power. max E represents the maximum reactive power output of the main topology of the T-type three-level converter when the voltage drops to its maximum allowable value. min The minimum allowable voltage amplitude when the main topology of a T-type three-level converter outputs maximum reactive power;

[0020] Construct the power outer loop; the specific formula for the power outer loop is as follows:

[0021]

[0022] In the formula, P is the active power output of the T-type three-level converter main topology when the frequency decreases, ω0 is the rated angular frequency output of the T-type three-level converter main topology, P0 is the rated active power output of the T-type three-level converter main topology, E0 is the rated voltage output of the T-type three-level converter main topology, Q0 is the rated reactive power output of the T-type three-level converter main topology, Q is the reactive power output of the T-type three-level converter main topology, and ω c G represents the minimum allowable voltage amplitude when the main topology of a T-type three-level converter outputs maximum reactive power. LP (s) is a filter used to remove higher harmonics in the calculation, where s is the complex frequency variable of the Laplace transform, and ω c The cutoff frequency of the filter is represented, where the droop equation includes the first and second lines of the above formula, and the power calculation equation includes the third, fourth, and fifth lines of the above formula.

[0023] Based on a typical Type I system, the proportional and integral coefficients of the current inner loop are obtained, and the power outer loop is obtained according to the droop formula, including:

[0024] The open-loop transfer function of the inner current loop in the S-domain is:

[0025]

[0026] In the formula, G oi (s) represents the open-loop transfer function value of the inner current loop, where s is the complex frequency variable of the Laplace transform, and K iPK represents the proportionality coefficient. PWM T represents the modulation gain. i =L / R represents the constant pole, T s The switching period is represented by L, the filter inductance value is represented by R, and the parasitic resistance of the LC filter inductor is represented by R.

[0027] Based on a typical Type I system, the proportional coefficient K of the inner current loop is obtained. ip and integral coefficient K iI :

[0028]

[0029] Prioritize step 2, performing a first-order active disturbance rejection paradigm transformation on the formula corresponding to the voltage loop in the mathematical model to obtain a second-order linear extended state observer, including:

[0030] The capacitor voltage state variable from step 1 is formalized into a first-order active disturbance rejection paradigm:

[0031]

[0032] In the formula, b0 = 1 / C f To control the gain, C f For the AC side LC filter, i 1d and i 1q These represent the d-axis and q-axis components of the filter inductor current in the main topology filter of a T-type three-level converter, respectively; f d f represents the total disturbance along the d-axis. q The total perturbation along the q-axis:

[0033]

[0034] Among them, w d The unknown disturbance representing the d-axis, w q The unknown perturbation represents the q-axis, while the coupling term is considered to be the known perturbation of the dq-axis; i 2d and i 2q These are the d-axis and q-axis components of the load current on the load side, respectively; u cd and u cq These are the d-axis and q-axis components of the output voltage, respectively.

[0035] Let x1 = u c Given x2 = f, the spatial state equation of the system is:

[0036]

[0037] In the formula, Let y be the derivative of the system state variable, y be the system output variable, and b0 be the control gain. For the disturbance differential term, u = i;

[0038] Since the gain of the inner current loop is often much greater than that of the outer current loop, therefore i ref =i;

[0039] A second-order linear extended state observer is established based on equation (7):

[0040]

[0041] In the formula, x1 and x2 are the estimated values ​​of the capacitor voltage and the total disturbance, respectively; β1 and β2 are the LESO feedback gain coefficients of the voltage loop; through pole placement, the observer feedback gain coefficients are configured within the observer bandwidth ω. o At this point, we get β1 = 2ω o ,

[0042]

[0043] Prioritize step 3, adding a perturbation differential term as a new state variable to the second-order linear extended observer obtained in step 2, constructing a new increased-order LESO with a perturbation differential term, and adding a hysteresis factor at the total perturbation output for initial noise suppression, including:

[0044] Expanding the differential term of the estimated total disturbance x2 into a new state variable x3 yields the corresponding observation. By observing the changing trend of the lumped perturbation, we obtain a new increased-order LESO with perturbation differential terms:

[0045]

[0046] Where β1=3ω o , b0 is the control gain, u = i;

[0047] Add a hysteresis factor to the total disturbance output side:

[0048]

[0049] In the formula, It is the lumped disturbance, T, after preliminary filtering by the hysteresis factor. e is the lag filter time factor, and s is the complex frequency variable of the Laplace transform.

[0050] Prioritizes step 4, by weighting and paralleling the second-order linear extended state observer obtained in step 2 with the increased-order LESO obtained in step 3, to obtain an adaptively weighted parallel LESO, including:

[0051] The second-order linear extended state observer obtained in step 2 is connected in weighted parallel with the increased-order LESO obtained in step 3 to obtain an adaptively weighted parallel LESO:

[0052]

[0053] In the formula, z1 is the estimated capacitor voltage output of the entire adaptive parallel weighted observer, and z2 is the estimated lumped disturbance output of the entire adaptive parallel weighted observer. It is the capacitor voltage estimated by the enhanced-order LESO method. It is the capacitor voltage estimated by a second-order linear extended state observer. It is the lumped perturbation estimated by the enhanced-order LESO after preliminary filtering with a lag factor. It is the lumped perturbation estimated by the second-order linear extended state observer, where δ is the adaptive weighting factor:

[0054]

[0055] In the formula, a, b, c, d, and f are all adjustment factors and are greater than 0, x ref x1 is the reference value for capacitor voltage, and x2 is the actual value of capacitor voltage.

[0056] Prioritize step 5 by introducing an improved adaptive superspiral sliding mode control law into the state error feedback control law. This law uses a hyperbolic tangent function to weaken the inherent chattering of the sliding mode and further enhances its disturbance rejection capability and transient performance based on the information from the lumped disturbance observations provided by the adaptively weighted parallel LESO. The disturbances in the mathematical modeling of step 2 include:

[0057] Design the sliding surface according to equation (7) in step 2:

[0058]

[0059] In the formula, s is the sliding surface function, c > 0 are the sliding integral coefficients, and x ref z1 is the capacitor voltage setpoint, z2 is the capacitor voltage estimate output by the adaptive weighted observer, and e is the voltage setpoint and estimation error.

[0060] When on the sliding surface, s = 0, let get:

[0061]

[0062] Replace the system state values ​​x1 and x2 in equation (7) of step 2 with the observed values ​​z1 and z2, and substitute them into equation (14) to obtain the equivalent control quantity u. eq :

[0063]

[0064] When outside the sliding surface, the approach rate For improved superhelical approach rate:

[0065]

[0066] In the formula, k1, k2, and k3 are the parameters to be designed for the sliding mode controller, all of which are greater than 0; r is the coefficient to be designed; tanh(s) is the hyperbolic tangent function; and sgn(s) is the sign function.

[0067] The above approach rate is adaptively adjusted so that the approach speed increases when the approach rate is far from the sliding surface and decreases when it is close to it:

[0068]

[0069] In the formula, a and b are adjustment factors;

[0070] Based on equations (14) and (17), the switching control quantity u is calculated. sw :

[0071] u sw =k1ρ1(|s|)tanh(s)+k2∫sgn(s)dt+k3ρ2(|e|)s(18),

[0072] Based on equation (18), the control u of the voltage sliding mode active disturbance rejection control output is calculated:

[0073] u = u eq +u sw (19).

[0074] Preferably, in step 6, based on the power outer loop, voltage loop, and current inner loop, the output modulation signal is obtained; the output modulation signal is transmitted to the SVPWM module with center-point balancing to control the main topology of the T-type three-level converter, including:

[0075] The control voltages calculated in steps 1 to 5 are sent to the inverse Park transform to obtain a three-phase voltage modulation wave. After per-unit processing, the modulation wave is sent to the SVPWM modulator and adjusted using the balance factor method based on the principle of midpoint charge conservation to output a PWM wave. The T-type converter is then controlled by 12 PWM control signals.

[0076] Preferably, the present invention provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the method described in any of the first aspects.

[0077] Preferably, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in any of the first aspects.

[0078] The beneficial effects achieved by this invention are as follows:

[0079] 1. In step 1, the power loop and current loop in the three-loop power, voltage and current system are initially constructed. Then, in step 2, a first-order active disturbance rejection control is built to improve its transient performance and disturbance rejection capability. In step 3, the disturbance infinitesimal component is added as a new state variable to build an upgraded order control, which further improves transient performance and disturbance rejection capability. The observer bandwidth is also increased by using the hysteresis factor.

[0080] 2. By adopting an adaptive parallel connection of enhanced LESO and traditional LESO in step 4, the proportion of enhanced LESO in steady state is reduced, while the proportion in transient and disturbance states is increased, so as to achieve low noise in steady state and good transient performance and disturbance rejection effect.

[0081] 3. By replacing the traditional linear control law with a sliding mode control law in step 5, the advantages of nonlinear control are brought into play, further improving its disturbance rejection performance and enhancing the robustness of the system. The disturbance differential increased-order LESO is fully utilized to improve the disturbance observation capability. The noise generated by the increased-order LESO is reduced by weighting it in parallel with the traditional LESO. Then, adaptive weighting is adopted to increase the proportion of the increased-order LESO during large disturbances and transient processes, thereby improving the control response speed and robustness. In steady state, the proportion of the traditional LESO is increased to reduce the impact of noise. This fully utilizes the advantages of the observer and compensates for its shortcomings.

[0082] 4. By adopting sliding mode control instead of traditional linear control, the advantages of nonlinear control are brought into play, further improving its disturbance rejection performance and enhancing the robustness of the system. Attached Figure Description

[0083] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0084] Figure 1 This is the overall control block diagram of the main topology of the T-type three-level converter proposed in this invention;

[0085] Figure 2 This is the main circuit topology diagram of the T-type three-level converter used in this invention;

[0086] Figure 3 This is a block diagram of the adaptive improved superspiral sliding mode control proposed in this invention;

[0087] Figure 4 This is the block diagram of the adaptive weighted parallel LESO control proposed in this invention;

[0088] Figure 5The diagram shows a comparison of the output voltage amplitudes of the improved adaptive sliding mode active disturbance rejection control and PI and LADRC control proposed in this invention under sudden load and voltage setpoint changes.

[0089] Figure 6 The diagram shows a comparison of the output active power of the improved adaptive sliding mode active disturbance rejection control and PI and LADRC control proposed in this invention under sudden load and voltage change.

[0090] Figure 7 This is a comparison of the output reactive power of the improved adaptive sliding mode active disturbance rejection control and PI and LADRC control proposed in this invention under sudden load and voltage change. Detailed Implementation

[0091] See Figure 1 This application discloses an off-grid control method for a T-type energy storage converter based on improved sliding mode self-disturbance rejection, including:

[0092] Step 1: Using the main topology of the T-type three-level converter, obtain the mathematical model in a two-phase rotating coordinate system; based on the typical type I system, obtain the proportional coefficient and integral coefficient of the current inner loop, and obtain the power outer loop according to the droop formula;

[0093] Step 2, according to Figure 3 and Figure 4 The structural design prepares to design a voltage loop, and transforms the formula corresponding to the voltage loop in the mathematical model into a first-order active disturbance rejection paradigm to obtain a second-order linear extended state observer.

[0094] Step 3: Add a perturbation differential term as a new state variable to the second-order linear extended observer obtained in Step 2, construct a new extended-order LESO with perturbation differential term, and add a hysteresis factor at the total perturbation output for initial noise suppression.

[0095] Step 4: Connect the second-order linear extended state observer obtained in Step 2 with the increased-order LESO obtained in Step 3 in a weighted parallel to obtain an adaptively weighted parallel LESO, which further reduces the noise amplification caused by the increased-order LESO. The weighting factor is also adaptively processed to increase the proportion of increased-order LESO during transient disturbances and decrease the proportion of increased-order LESO during steady state.

[0096] Step 5: An improved adaptive super-twisting sliding mode (STSM) control law is introduced into the state error feedback control law (SEF). The hyperbolic tangent function is used to weaken the inherent chattering of the sliding mode. Based on the information of the lumped disturbance observations provided by the adaptive weighted parallel LESO, a voltage loop is constructed to suppress the disturbances in the mathematical modeling of the active disturbance rejection paradigm in Step 2, so as to achieve fast and accurate control of the system state.

[0097] Step 6: Adjust the power outer loop, voltage loop, and current inner loop to obtain the output modulation signal; transmit the output modulation signal to the SVPWM module with center point balancing to control the main topology of the T-type three-level converter.

[0098] Step 1 includes:

[0099] Using the main topology of a T-type three-level converter, a mathematical model in a two-phase rotating coordinate system is obtained:

[0100]

[0101] In the formula, C f For the AC side LC filter, L is the filter capacitor. f For the filter inductance of the LC filter, i 1d and i 1q These represent the d-axis and q-axis components of the filter inductor current in the main topology filter of a T-type three-level converter, respectively; 2d and i 2q These are the d-axis and q-axis components of the load current on the load side, respectively; u cd and u cq These are the d-axis and q-axis components of the output voltage, respectively. d For the main topology of a T-type three-level converter, the output d-axis voltage, u q The output q-axis voltage of the main topology of the T-type three-level converter is given by R1, which is the parasitic resistance of the filter inductor of the AC side LC filter, and ω is the known angular frequency corresponding to the power frequency. The formula corresponding to the voltage loop is the third and fourth formulas from top to bottom in formula (1).

[0102] According to Article 3.1 of the national standard "Power Quality - Power System Frequency Amplitude Deviation", which stipulates that the voltage angular frequency variation of the power grid should be less than or equal to 1% and the amplitude variation should be less than or equal to 5%, the droop factor is determined as follows:

[0103]

[0104] In the formula, P max Let ω0 be the maximum active power output of the T-type three-level converter main topology when the frequency decreases, P0 be the rated active power output of the T-type three-level converter main topology, E0 be the rated voltage output of the T-type three-level converter main topology, and Q0 be the rated reactive power output of the T-type three-level converter main topology. min Q is the minimum permissible angular frequency (Q) for the main topology of a T-type three-level converter to output maximum active power. maxE represents the maximum reactive power output of the main topology of the T-type three-level converter when the voltage drops to its maximum allowable value. min The minimum allowable voltage amplitude when the main topology of a T-type three-level converter outputs maximum reactive power;

[0105] Construct the power outer loop; the specific formula for the power outer loop is as follows:

[0106]

[0107] In the formula, P is the active power output of the T-type three-level converter main topology when the frequency decreases, ω0 is the rated angular frequency output of the T-type three-level converter main topology, P0 is the rated active power output of the T-type three-level converter main topology, E0 is the rated voltage output of the T-type three-level converter main topology, Q0 is the rated reactive power output of the T-type three-level converter main topology, Q is the reactive power output of the T-type three-level converter main topology, and ω c G represents the minimum allowable voltage amplitude when the main topology of a T-type three-level converter outputs maximum reactive power. LP (s) is a filter used to remove higher harmonics in the calculation, where s is the complex frequency variable of the Laplace transform, and ω c The cutoff frequency of the filter is represented, where the droop equation includes the first and second lines of the above formula, and the power calculation equation includes the third, fourth, and fifth lines of the above formula.

[0108] Based on a typical Type I system, the proportional and integral coefficients of the current inner loop are obtained, and the power outer loop is obtained according to the droop formula, including:

[0109] The open-loop transfer function of the inner current loop in the S-domain is:

[0110]

[0111] In the formula, G oi (s) represents the open-loop transfer function value of the inner current loop, where s is the complex frequency variable of the Laplace transform, and K iP K represents the proportionality coefficient. PWM T represents the modulation gain. i =L / R represents the constant pole, T s The switching period is represented by L, the filter inductance value is represented by R, and the parasitic resistance of the LC filter inductor is represented by R.

[0112] Based on a typical Type I system, the proportional coefficient K of the inner current loop is obtained. ip and integral coefficient K iI :

[0113]

[0114] Step 2, according to Figure 3 and Figure 4 The structural design transforms the formula corresponding to the voltage loop in the mathematical model using a first-order active disturbance rejection paradigm to obtain a second-order linear extended state observer, including:

[0115] The capacitor voltage state variable from step 1 is formalized into a first-order active disturbance rejection paradigm:

[0116]

[0117] In the formula, b0 = 1 / C f To control the gain, C f For the AC side LC filter, i 1d and i 1q These represent the d-axis and q-axis components of the filter inductor current in the main topology filter of a T-type three-level converter, respectively; f d f represents the total disturbance along the d-axis. q The total perturbation along the q-axis:

[0118]

[0119] Among them, w d The unknown disturbance representing the d-axis, w q The unknown perturbation represents the q-axis, while the coupling term is considered to be the known perturbation of the dq-axis; i 2d and i 2q These are the d-axis and q-axis components of the load current on the load side, respectively; u cd and u cq These are the d-axis and q-axis components of the output voltage, respectively; therefore, the d and q components have the same structure and can be designed separately, so they will not be specifically distinguished below.

[0120] Let x1 = u c Given x2 = f, the spatial state equation of the system is:

[0121]

[0122] In the formula, Let y be the derivative of the system state variable, y be the system output variable, and b0 be the control gain. For the disturbance differential term, u = i;

[0123] Since the gain of the inner current loop is often much greater than that of the outer current loop, i can be considered... ref =i;

[0124] A second-order linear extended state observer is established based on equation (7):

[0125]

[0126] In the formula, x1 and x2 are the estimated values ​​of the capacitor voltage and the total disturbance, respectively; β1 and β2 are the LESO feedback gain coefficients of the voltage loop; through pole placement, the observer feedback gain coefficients are configured within the observer bandwidth ω. o At this point, we get β1 = 2ω o ,

[0127]

[0128] Step 3: Add a perturbation differential term as a new state variable to the second-order linear extended observer obtained in Step 2, construct a new increased-order LESO with a perturbation differential term, and add a hysteresis factor at the total perturbation output for initial noise suppression, including:

[0129] To further improve the observation performance and control bandwidth of the LESO observer, the disturbance differential term of the estimated total disturbance x2 is expanded into a new state variable x3, yielding the corresponding observation value. By observing the changing trend of the lumped perturbation, we obtain a new increased-order LESO with perturbation differential terms:

[0130]

[0131] Where β1=3ω o , b0 is the control gain, u = i;

[0132] To broaden LESO's estimation range for total disturbances, it can effectively address the issue of excessively large ω. ο The resulting system instability is addressed by adding a hysteresis factor to the total disturbance output side:

[0133]

[0134] In the formula, It is the lumped disturbance, T, after preliminary filtering by the hysteresis factor. e is the lag filter time factor, and s is the complex frequency variable of the Laplace transform.

[0135] Step 4: Connect the second-order linear extended state observer obtained in Step 2 in a weighted parallel with the increased-order LESO obtained in Step 3 to obtain an adaptively weighted parallel LESO, including:

[0136] To further address the noise amplification caused by the increased-order differential term, the second-order linear extended state observer obtained in step 2 is connected in weighted parallel with the increased-order LESO obtained in step 3 to obtain an adaptively weighted parallel LESO:

[0137]

[0138] In the formula, z1 is the estimated capacitor voltage output of the entire adaptive parallel weighted observer, and z2 is the estimated lumped disturbance output of the entire adaptive parallel weighted observer. It is the capacitor voltage estimated by the enhanced-order LESO method. It is the capacitor voltage estimated by a second-order linear extended state observer. It is the lumped perturbation estimated by the enhanced-order LESO after preliminary filtering with a lag factor. It is the lumped perturbation estimated by the second-order linear extended state observer, where δ is the adaptive weighting factor:

[0139]

[0140] In the formula, a, b, c, d, and f are all adjustment factors and are greater than 0, x ref x1 is the reference value of the capacitor voltage and x2 is the actual value of the capacitor voltage. With this adaptive factor δ, the proportion of the increased-order LESO is about 0.5 when in steady state, and it rapidly increases to 0.8 when in a large disturbance or transient process.

[0141] Step 5 introduces an improved adaptive superspiral sliding mode control law into the state error feedback control law. The hyperbolic tangent function is used to weaken the inherent chattering of the sliding mode. Furthermore, based on the information from the lumped disturbance observations provided by the adaptively weighted parallel LESO, its disturbance rejection capability and transient performance are further enhanced. The disturbances in the mathematical modeling of Step 2 include:

[0142] Design the sliding surface according to equation (7) in step 2:

[0143]

[0144] In the formula, s is the sliding surface function, c > 0 are the sliding integral coefficients, and x ref z1 is the capacitor voltage setpoint, z2 is the capacitor voltage estimate output by the adaptive weighted observer, and e is the voltage setpoint and estimation error.

[0145] When on the sliding surface, s = 0, let get:

[0146]

[0147] Replace the system state values ​​x1 and x2 in equation (7) of step 2 with the observed values ​​z1 and z2, and substitute them into equation (14) to obtain the equivalent control quantity u. eq :

[0148]

[0149] When outside the sliding surface, the approach rate For improved superhelical approach rate:

[0150]

[0151] In the formula, k1, k2, and k3 are the parameters to be designed for the sliding mode controller, all of which are greater than 0; r is the coefficient to be designed; tanh(s) is the hyperbolic tangent function; and sgn(s) is the sign function.

[0152] The above approach rate is adaptively adjusted so that the approach speed increases when the approach rate is far from the sliding surface and decreases when it is close to it:

[0153]

[0154] In the formula, a and b are adjustment factors;

[0155] Based on equations (14) and (17), the switching control quantity u is calculated. sw :

[0156] u sw =k1ρ1(|s|)tanh(s)+k2∫sgn(s)dt+k3ρ2(|e|)s(18),

[0157] Based on equation (18), the control u of the voltage sliding mode active disturbance rejection control output is calculated:

[0158] u = u eq +u sw (19).

[0159] In step 6, the power, voltage, and current three-loop control outputs are sent to the SVPWM module with center-point balancing to control the main topology of the T-type three-level converter, including:

[0160] The control voltages calculated in steps 1 to 5 are sent to the inverse Park transform to obtain a three-phase voltage modulation wave. After per-unit processing, the modulation wave is sent to the SVPWM modulator and adjusted using the balance factor method based on the principle of midpoint charge conservation to output a PWM wave. The overall control strategy is realized by controlling the operation of the T-type converter through 12 PWM control signals.

[0161] The control voltage output from the power-voltage-current three-loop design in steps 1 to 5 is fed to the inverse Park converter to obtain a three-phase voltage modulation wave. After per-unit processing, the modulation wave is fed into the SVPWM modulator and adjusted using the balance factor method based on the midpoint charge conservation principle to output a PWM wave. The overall control strategy is achieved by controlling the T-type converter through 12 PWM control signals. Figure 1 As shown, the present invention includes a T-type energy storage converter, an LC filter, and a first line impedance R. C Second line impedance L C ,

[0162] The T-type energy storage converter includes an IGBT and a diode. A diode is connected in reverse parallel to the collector (C) and emitter (E) terminals of the IGBT. The specific structure is as follows: Figure 2 As shown, the current i1 output by the T-type energy storage converter is sequentially input to the LC filter and the first line impedance R. c Second line impedance L c and load, based on the filter capacitor C of the LC filter f voltage v cabc and the output current i flowing through the line impedance 2abc Perform power calculations and add the filter capacitor C of the LC filter. f voltage v cabc Input to the adaptive weighted parallel LESO; output current i 2abc And the filter capacitor C of the LC filter f The voltage is transformed by dq and then used for power calculation. The output active power P and reactive power Q are compared with the rated active power P0, rated reactive power Q0, rated frequency ω0, and rated voltage E. o Input the droop equation, and use the parameters ω and E output by the droop equation to generate a reference voltage, thus obtaining V. oref ; For the adaptive weighted parallel LESO outputs Z1 and Z2, V oref Adaptive improvement of superspiral sliding mode processing is performed, and the adaptive improvement of superspiral sliding mode output current i is achieved. ref and current i labc After current PI control, the output is sent to SVPWM.

[0163] Figure 2 The main circuit topology diagram of the T-type three-level converter used in this invention includes:

[0164] Power supply U dc The first capacitor C1 and the second capacitor C2 are connected in series, and the power supply U dc Upper, first insulated gate bipolar transistor Q a1 The third insulated gate bipolar transistor Q a3 The second insulated gate bipolar transistor Q a2 The second capacitor C2 is connected in series, and the first insulated gate bipolar transistor Q... a1 The first diode D connected in anti-parallel a1 The second insulated-gate bipolar transistor Q a2 The second diode D is connected in anti-parallel. a2 The third insulated-gate bipolar transistor Q a3 The third diode D is connected in anti-parallel. a3 Power supply U dc Upper, fourth insulated gate bipolar transistor Q b1 The sixth insulated-gate bipolar transistor Q b3Fifth Insulated Gate Bipolar Transistor Q b2 The second capacitor C2 is connected in series, and the fourth insulated gate bipolar transistor Q... b1 The fourth diode D is connected in anti-parallel. b1 The fifth insulated-gate bipolar transistor Q b2 The fifth diode D is connected in anti-parallel. b2 The sixth insulated-gate bipolar transistor Q b3 The sixth diode D is connected in anti-parallel. b3 Power supply U dc Upper, seventh insulated-gate bipolar transistor Q c1 The ninth insulated-gate bipolar transistor Q c3 The eighth insulated-gate bipolar transistor Q c2 The second capacitor C2 is connected in series, and the seventh insulated-gate bipolar transistor Q... c1 The seventh diode D is connected in anti-parallel. c1 The eighth insulated-gate bipolar transistor Q c2 The eighth diode D is connected in anti-parallel. c2 The ninth insulated-gate bipolar transistor Q c3 The ninth diode D is connected in anti-parallel. c3 The first insulated gate bipolar transistor Q a1 The third insulated gate bipolar transistor Q a3 Voltage output connection point u between a and power supply U dc The fourth insulated-gate bipolar transistor Q is connected in series at the connection point N between the second capacitor C2 and the second capacitor C2. a4 The tenth insulated-gate bipolar transistor Q a4 The tenth diode D is connected in anti-parallel. a4 The fourth insulated-gate bipolar transistor Q b1 The sixth insulated-gate bipolar transistor Q b3 Voltage output connection point u between b and power supply U dc The eleventh insulated-gate bipolar transistor Q is connected in series at the connection point N between the second capacitor C2 and the second capacitor C2. b4 Eleventh Insulated Gate Bipolar Transistor Q b4 There is an eleventh diode D connected in anti-parallel. b4 The seventh insulated-gate bipolar transistor Q c1 The ninth insulated-gate bipolar transistor Q c3 Voltage output connection point u between c and power supply U dc The twelfth insulated-gate bipolar transistor Q is connected in series at the connection point N between the second capacitor C2 and the second capacitor C2. c4 The twelfth insulated-gate bipolar transistor Q c4 The twelfth diode D is connected in anti-parallel. c4 Voltage output connection point ua u b u c The filter inductor L of the LC filter is connected to the filter. f parasitic resistance R of filter inductor f The filter capacitor C of the LC filter is Y-connected between the three lines. f Intersection at n c Point, the connection point u of three pairs of filter inductors and filter capacitors. ca u cb u cc Connect the load.

[0165] The invention and its reverse verification were performed in Matlab / Simulink. A simulation model was built based on the derivation of the above steps. Initially, a 15kW+9kVar load was used. At 0.4s, a 15kW+9kVar load was switched in again. At 0.8s, the voltage setpoint dropped by 50V. At 1s, the voltage setpoint was restored.

[0166] Figure 5 This is a comparison of the output voltage amplitude of the improved adaptive sliding mode active disturbance rejection control and PI and LADRC control proposed in this invention under sudden load and voltage setpoint changes. The improved scheme of this invention has a smooth voltage amplitude change during the startup process without sudden drop, and a short adjustment time. When the load is switched on, the voltage drop is small and the recovery speed is fast. When the voltage setpoint changes, it tracks the fastest and the voltage amplitude ripple change is smaller in each steady state stage.

[0167] Figure 6 This is a comparison chart of the output active power of the improved adaptive sliding mode active disturbance rejection control and PI and LADRC control proposed in this invention under sudden load and voltage setpoint changes. Figure 7 The above two figures show a comparison of the output reactive power of the improved adaptive sliding mode active disturbance rejection control and PI / LADRC control proposed in this invention under sudden load and voltage setpoint changes. As can be seen from the figures, the improved scheme of this invention has the shortest rise time and the smoothest transition during both the startup and disturbance occurrence processes.

[0168] In this embodiment of the application, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of any of the methods described above.

[0169] In this application embodiment, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0170] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0171] Other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention described herein. This application is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not invented herein. The specification and embodiments are to be considered exemplary only.

[0172] The above specific embodiments further illustrate the purpose, technical solution and beneficial effects of this application. It should be understood that the above are only specific embodiments of this application and are not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made on the basis of the technical solution of this application should be included within the scope of protection of this application.

Claims

1. An off-grid control method for T-type energy storage converters based on improved sliding mode self-disturbance rejection, characterized in that, include: Step 1: Using the main topology of the T-type three-level converter, obtain the mathematical model in a two-phase rotating coordinate system; The proportional and integral coefficients of the current inner loop are obtained based on a typical Type I system, and the power outer loop is obtained according to the droop formula. Step 2: Transform the formula corresponding to the voltage loop in the mathematical model into the first-order active disturbance rejection paradigm to obtain the second-order linear extended state observer. Step 3: Add a perturbation differential term as a new state variable to the second-order linear extended observer obtained in Step 2, construct a new extended-order LESO with perturbation differential term, and add a hysteresis factor at the total perturbation output for initial noise suppression. Step 4: Connect the second-order linear extended state observer obtained in Step 2 with the increased-order LESO obtained in Step 3 in a weighted parallel to obtain an adaptively weighted parallel LESO. Step 5: An improved adaptive superspiral sliding mode control law is introduced into the state error feedback control law. The hyperbolic tangent function is used to weaken the inherent chattering of the sliding mode, and a voltage loop is constructed based on the information of the lumped disturbance observations provided by the adaptive weighted parallel LESO. Step 6: Based on the power outer loop, voltage loop, and current inner loop, obtain the output modulation signal; transmit the output modulation signal to the SVPWM module with center point balancing to control the main topology of the T-type three-level converter.

2. The off-grid control method for T-type energy storage converter based on improved sliding mode self-disturbance rejection as described in claim 1, characterized in that, Step 1 includes: Using the main topology of a T-type three-level converter, a mathematical model in a two-phase rotating coordinate system is obtained: In the formula, C f For the AC side LC filter, L is the filter capacitor. f For the filter inductance of the LC filter, i 1d and i 1q These represent the d-axis and q-axis components of the filter inductor current in the main topology filter of a T-type three-level converter, respectively; 2d and i 2q These are the d-axis and q-axis components of the load current on the load side, respectively; u cd and u cq These are the d-axis and q-axis components of the output voltage, respectively. d For the main topology of a T-type three-level converter, the output d-axis voltage, u q The output q-axis voltage of the main topology of the T-type three-level converter is given by R1, which is the parasitic resistance of the filter inductor of the AC side LC filter, and ω is the known angular frequency corresponding to the power frequency. The formula corresponding to the voltage loop is the third and fourth formulas from top to bottom in formula (1). Based on the requirement that the voltage angular frequency variation of the power grid should be less than or equal to 1% and the amplitude variation should be less than or equal to 5%, the droop factor is determined as follows: In the formula, P max Let ω0 be the maximum active power output of the T-type three-level converter main topology when the frequency decreases, P0 be the rated active power output of the T-type three-level converter main topology, E0 be the rated voltage output of the T-type three-level converter main topology, and Q0 be the rated reactive power output of the T-type three-level converter main topology. min Q is the minimum permissible angular frequency (Q) for the main topology of a T-type three-level converter to output maximum active power. max E represents the maximum reactive power output of the main topology of the T-type three-level converter when the voltage drops to its maximum allowable value. min The minimum allowable voltage amplitude when the main topology of a T-type three-level converter outputs maximum reactive power; Construct the power outer loop; the specific formula for the power outer loop is as follows: In the formula, P is the active power output of the T-type three-level converter main topology when the frequency decreases, ω0 is the rated angular frequency output of the T-type three-level converter main topology, P0 is the rated active power output of the T-type three-level converter main topology, E0 is the rated voltage output of the T-type three-level converter main topology, Q0 is the rated reactive power output of the T-type three-level converter main topology, Q is the reactive power output of the T-type three-level converter main topology, and ω c G represents the minimum allowable voltage amplitude when the main topology of a T-type three-level converter outputs maximum reactive power. LP (s) is a filter used to remove higher harmonics in the calculation, where s is the complex frequency variable of the Laplace transform, and ω c The cutoff frequency of the filter is represented, where the droop equation includes the first and second lines of the above formula, and the power calculation equation includes the third, fourth, and fifth lines of the above formula. Based on a typical Type I system, the proportional and integral coefficients of the current inner loop are obtained, and the power outer loop is obtained according to the droop formula, including: The open-loop transfer function of the inner current loop in the S-domain is: In the formula, G oi (s) represents the open-loop transfer function value of the inner current loop, where s is the complex frequency variable of the Laplace transform, and K iP K represents the proportionality coefficient. PWM T represents the modulation gain. i =L / R represents the constant pole, T s The switching period is represented by L, the filter inductance value is represented by R, and the parasitic resistance of the LC filter inductor is represented by R. Based on a typical Type I system, the proportional coefficient K of the inner current loop is obtained. ip and integral coefficient K iI :

3. The off-grid control method for T-type energy storage converter based on improved sliding mode self-disturbance rejection as described in claim 1, characterized in that, Step 2: Perform a first-order active disturbance rejection paradigm transformation on the formula corresponding to the voltage loop in the mathematical model to obtain a second-order linear extended state observer, including: The capacitor voltage state variable from step 1 is formalized into a first-order active disturbance rejection paradigm: In the formula, b0 = 1 / C f To control the gain, C f For the AC side LC filter, i 1d and i 1q These represent the d-axis and q-axis components of the filter inductor current in the main topology filter of a T-type three-level converter, respectively; f d f represents the total disturbance along the d-axis. q The total perturbation along the q-axis: Among them, w d The unknown disturbance representing the d-axis, w q The unknown perturbation represents the q-axis, and the coupling term is considered as the known perturbation of the dq-axis; i 2d and i 2q These are the d-axis and q-axis components of the load current on the load side, respectively; u cd and u cq These are the d-axis and q-axis components of the output voltage, respectively. Let x1 = u c Given x2 = f, the spatial state equation of the system is: In the formula, and Let y be the derivative of the system state variable, y be the system output variable, and b0 be the control gain. For the disturbance differential term, u = i; Since the gain of the inner current loop is often much greater than that of the outer current loop, therefore i ref =i; A second-order linear extended state observer is established based on equation (7): In the formula, x1 and x2 are the estimated values ​​of the capacitor voltage and the total disturbance, respectively; β1 and β2 are the LESO feedback gain coefficients of the voltage loop; through pole placement, the observer feedback gain coefficients are configured within the observer bandwidth ω. o At this point, we get β1 = 2ω o , 4. The off-grid control method for T-type energy storage converter based on improved sliding mode self-disturbance rejection as described in claim 3, characterized in that, Step 3: Add a perturbation differential term as a new state variable to the second-order linear extended observer obtained in Step 2, construct a new increased-order LESO with a perturbation differential term, and add a hysteresis factor at the total perturbation output for initial noise suppression, including: Expanding the differential term of the estimated total disturbance x2 into a new state variable x3 yields the corresponding observation. By observing the changing trend of the lumped perturbation, we obtain a new increased-order LESO with perturbation differential terms: Where β1'=3ω o , b0 is the control gain, u = i; Add a hysteresis factor to the total disturbance output side: In the formula, It is the lumped disturbance, T, after preliminary filtering by the hysteresis factor. e is the lag filter time factor, and s is the complex frequency variable of the Laplace transform.

5. The off-grid control method for a T-type energy storage converter based on improved sliding mode self-disturbance rejection as described in claim 1, characterized in that, Step 4: Connect the second-order linear extended state observer obtained in Step 2 in a weighted parallel with the increased-order LESO obtained in Step 3 to obtain an adaptively weighted parallel LESO, including: The second-order linear extended state observer obtained in step 2 is connected in weighted parallel with the increased-order LESO obtained in step 3 to obtain an adaptively weighted parallel LESO: In the formula, z1 is the estimated capacitor voltage output of the entire adaptive parallel weighted observer, and z2 is the estimated lumped disturbance output of the entire adaptive parallel weighted observer. It is the capacitor voltage estimated by the enhanced-order LESO method. It is the capacitor voltage estimated by a second-order linear extended state observer. It is the lumped perturbation estimated by the enhanced-order LESO after preliminary filtering with a lag factor. It is the lumped perturbation estimated by the second-order linear extended state observer, where δ is the adaptive weighting factor: In the formula, a, b, c, d, and f are all adjustment factors and are greater than 0 and x. ref x1 is the reference value for capacitor voltage, and x2 is the actual value of capacitor voltage.

6. The off-grid control method for a T-type energy storage converter based on improved sliding mode self-disturbance rejection as described in claim 3, characterized in that, Step 5 introduces an improved adaptive superspiral sliding mode control law into the state error feedback control law. The hyperbolic tangent function is used to weaken the inherent chattering of the sliding mode. Furthermore, based on the information from the lumped disturbance observations provided by the adaptively weighted parallel LESO, its disturbance rejection capability and transient performance are further enhanced. The disturbances in the mathematical modeling of Step 2 include: Design the sliding surface according to equation (7) in step 2: In the formula, s is the sliding surface function, c > 0 are the sliding integral coefficients, and x ref z1 is the capacitor voltage setpoint, z2 is the capacitor voltage estimate output by the adaptive weighted observer, and e is the voltage setpoint and estimation error. When on the sliding surface, s = 0, let get: Replace the system state values ​​x1 and x2 in equation (7) of step 2 with the observed values ​​z1 and z2, and substitute them into equation (14) to obtain the equivalent control quantity u. eq : When outside the sliding surface, the approach rate s is the improved superspiral approach rate: In the formula, k1, k2, and k3 are the parameters to be designed for the sliding mode controller, all of which are greater than 0; r is the coefficient to be designed; tanh(s) is the hyperbolic tangent function; and sgn(s) is the sign function. The above approach rate is adaptively adjusted so that the approach speed increases when the approach rate is far from the sliding surface and decreases when it is close to it: In the formula, a and b are adjustment factors; Based on equations (14) and (17), the switching control quantity u is calculated. sw : u sw =k1ρ1(|s|)tanh(s)+k2∫sgn(s)dt+k3ρ2(|e|)s (18), Based on equation (18), the control u of the voltage sliding mode active disturbance rejection control output is calculated: in=in eq +in sw (19).

7. The off-grid control method for a T-type energy storage converter based on improved sliding mode self-disturbance rejection as described in claim 1, characterized in that, Step 6 involves obtaining the output modulation signal based on the power outer loop, voltage loop, and current inner loop; transmitting the output modulation signal to the SVPWM module with center-point balancing to control the main topology of the T-type three-level converter, including: The control voltages calculated in steps 1 to 5 are sent to the inverse Park transform to obtain a three-phase voltage modulation wave. After per-unit processing, the modulation wave is sent to the SVPWM modulator and adjusted using the balance factor method based on the principle of midpoint charge conservation to output a PWM wave. The T-type converter is then controlled by 12 PWM control signals.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method according to any one of claims 1 to 7.

Images