A deadbeat predictive current control method based on a quasi-proportional-resonant state observer

Abstract

This invention discloses a deadbeat predictive current control method based on a quasi-proportional resonant state observer, comprising: 1) establishing an accurate discrete model of a permanent magnet synchronous motor in a two-phase stationary coordinate system, describing the continuous changes of current, rotor electrical angle, and back electromotive force within a sampling period through analytical discretization; 2) extending the total disturbance caused by parameter mismatch, system uncertainty, sampling delay, and low carrier ratio operation into an observation state, constructing a quasi-proportional resonant extended state observer, and estimating and compensating for DC and AC disturbance components online; 3) the controller calculates the two-phase stationary coordinate axis command voltage for the next control period based on the accurate discrete model, one-step delay compensation, and disturbance estimation value from the quasi-proportional resonant extended state observer. This invention can improve current prediction accuracy and disturbance suppression capability under low carrier ratio and motor parameter mismatch conditions, and improve the current tracking performance and robustness of permanent magnet synchronous motors.

Description

Technical Field

[0001] This invention relates to the field of motor drive control technology, and in particular to a deadbeat predictive current control method based on a quasi-proportional resonant state observer. Background Technology

[0002] Permanent magnet synchronous motors (PMSMs) are widely used in electric vehicles, high-speed electric spindles, aerospace electric drives, servo systems, and high-performance industrial drives due to their high power density, high efficiency, and fast torque response. The performance of the current loop directly affects the system's torque response, steady-state accuracy, and robustness. Traditional PI control structures are simple, but they struggle to balance dynamic and steady-state performance under high-speed, high-power, or low carrier ratio conditions. While deadbeat predictive current control can quickly calculate the control voltage based on a discrete model and achieve rapid current tracking, at low carrier ratios, the forward Euler discretization error increases, parameter mismatch easily leads to steady-state current errors, and AC and time-varying disturbances are difficult to compensate effectively. Therefore, it is necessary to study novel predictive current control methods suitable for low carrier ratios and parameter mismatch conditions to improve current prediction accuracy, steady-state tracking performance, and disturbance suppression capabilities.

[0003] To achieve deadbeat-free predictive control of permanent magnet synchronous motors (PMSMs) under low carrier ratio conditions, scholars both domestically and internationally have conducted relevant research. Chinese invention patent CN117424512A discloses a method for accurate deadbeat-free predictive current control of PMSMs under low carrier ratio conditions. It derives precise determinant voltage equations for the synchronous rotating coordinate system of the PMSM rotor considering rotor orientation errors, and establishes current prediction and command voltage calculation models for the synchronous rotating coordinate system of the PMSM rotor considering rotor orientation errors. This solves the problem of direct-axis coupling caused by rotor flux orientation errors under low carrier ratio conditions, eliminates prediction errors and command voltage errors caused by multi-coordinate system variable cross-space calculations, and improves the torque flux decoupling capability, dq-axis current prediction accuracy, and tracking accuracy of PMSMs under low carrier ratio conditions. Chinese invention patent CN111641363A discloses a deadbeat control method for permanent magnet synchronous motors (PMSMs) under low carrier ratios. It employs an accurate discrete model of the dq-axis, designs a deadbeat current controller for low carrier ratios, and utilizes a hybrid modulation scheme of SVPWM and SHEPWM. SVPWM modulation is used at high carrier ratios, while SHEPWM modulation is used at low carrier ratios. This solves the problem of increased motor model error caused by using the forward Euler approximation at low carrier ratios, improves current control accuracy, compensates for delays at low carrier ratios, and effectively improves the voltage harmonic performance of the inverter and PMSM under low carrier ratios. However, deadbeat predictive control is more sensitive to parameters, making control more complex under low carrier ratio conditions. Therefore, how to achieve real-time observation of the system current of the PMSM under parameter changes during low carrier ratio conditions, compensate for disturbances caused by AC disturbances, parameter changes, and unmodeled dynamics, and effectively suppress harmonic currents are urgent technical problems to be solved. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a deadbeat predictive current control method based on a quasi-proportional resonant state observer. This method enables effective observation of system current and AC disturbances of a permanent magnet synchronous motor under parameter variations, compensates for disturbances caused by AC disturbances, parameter variations, and unmodeled dynamics, and reduces system chattering, thereby improving control robustness under low carrier ratio conditions.

[0005] The objective of this invention is achieved as follows: a deadbeat predictive current control method based on a quasi-proportional resonant state observer, comprising the following steps:

[0006] 1) Collect the three-phase stator current, rotor electrical angle and rotor electrical angular velocity of the permanent magnet synchronous motor, and transform the three-phase stator current to a two-phase stationary coordinate system to obtain the α-axis current and β-axis current;

[0007] 2) In a stationary two-phase stationary coordinate system, an accurate discrete current prediction model is established based on the continuous-time voltage equation of the permanent magnet synchronous motor;

[0008] 3) Construct a quasi-proportional resonance extended state observer (QR-ESO) to observe the total disturbance caused by motor parameter mismatch, model uncertainty, sampling delay, and low carrier ratio operation as an extended state online, and obtain the αβ axis disturbance estimate; perform one-step current prediction based on the accurate discrete current prediction model, and perform delay compensation in combination with the one-step delay characteristics of the digital control system;

[0009] 4) The αβ axis reference voltage, accurately predicted by the voltage vector, is input to the space vector pulse width modulated (PWM). The output PWM wave is processed by the inverter to obtain the drive voltage, which drives the permanent magnet synchronous motor. The actual speed of the permanent magnet synchronous motor during operation is detected by a position sensor. The difference between the actual speed and the given speed is then used by PI control to output the reference speed i. q Value, and reference i d The values ​​are transformed by coordinates to obtain the αβ axis reference current for voltage vector prediction, forming a speed closed-loop control system.

[0010] Furthermore, step 2) specifically includes establishing an accurate discrete current prediction model based on the continuous-time voltage equation of the permanent magnet synchronous motor, which specifically includes:

[0011] The precise discrete current prediction model is established in a stationary αβ coordinate system, and its form is as follows:

[0012]

[0013]

[0014] In the formula, i α (k), u α (k) represent the α-axis current and α-axis voltage at time k, respectively. β (k), u β (k) represents the β-axis current and β-axis voltage at time k, A is the current state transition matrix, B is the voltage input matrix, G(k) is the precisely discrete term of the back electromotive force related to the permanent magnet flux linkage, rotor electric angular velocity, and rotor electric angle, and L s For stator inductance; R is the sampling period; s Stator resistance; ψ f For rotor flux linkage, ω e (k) represents the electrical angular velocity at time k;

[0015] The exact discrete current prediction model is rewritten as a voltage equation, and a one-step lead-delay compensation is used to offset the delay introduced by digital control, resulting in the following equation:

[0016]

[0017] Among them, u α (k+1) represents the predicted value of the α-axis voltage, u β (k+1) represents the predicted value of the β-axis voltage, i α (k+1) represents the predicted value of the α-axis current, i β (k+1) represents the predicted value of the β-axis current, i α (k+2) is the reference value for the α-axis current, i β (k+2) is the reference value of the β-axis current, and G[k+1] is the precise discrete term of the back electromotive force at time k+1 related to the permanent magnet flux linkage, rotor electric angular velocity and rotor electric angle.

[0018] Furthermore, step 3) specifically includes constructing the quasi-proportional resonance extended state observer QR-ESO:

[0019]

[0020] Where x1 is the observed system output, and x2 is the total disturbance. and x1 and x2 are the observed values, respectively; β1 and β2 are the gains of the quasi-proportional resonant extended state observer QR-ESO, expressed as β1 = ξω0 and β2 = ω0. 2 ξ is the tuning coefficient that adjusts the dynamic response, ω0 is the bandwidth of the quasi-proportional resonant extended state observer QR-ESO; G(s) is the transfer function of the quasi-resonant controller.

[0021]

[0022] In the formula, k r For the resonant gain, ω r For the selected resonant frequency, ω c The cutoff frequency is given; the transfer function of the quasi-proportional resonant extended state observer QR-ESO is:

[0023] ;

[0024] Furthermore, the α-axis and β-axis observers of the quasi-proportional resonant extended state observer QR-ESO have the same structure. The α-axis observer introduces the lumped disturbance as an extended state into the α-axis current equation, thus obtaining the α-axis permanent magnet synchronous motor model for observer design:

[0025]

[0026] In the formula, i α F is the α-axis current value. αThis represents the lumped disturbance value along the α-axis. and is i α and F α The estimated value;

[0027] The above formula can be further rewritten as:

[0028]

[0029] in, The DC component is caused by parameter mismatch. AC component caused by low carrier ratio operation

[0030] Discretizing the aforementioned observer model, the discrete quasi-proportional resonant extended state observer QR-ESO is expressed as follows:

[0031]

[0032] By adjusting the observer parameters, the total α-axis disturbance F is estimated. α (k+1), thus obtaining the α-axis current i α The current equation is rewritten in voltage form, and a one-step delay compensation e is introduced. α Then, the equation at time k+1 is expressed as:

[0033] .

[0034] Furthermore, the stability of the discrete quasi-proportional resonant extended state observer QR-ESO is determined based on the Jury stability criterion, selecting the α-axis current i. α Total disturbance F along the α axis α Using the quasi-resonant controller state variables x1 and x2 as state variables, the discrete formula for QR-ESO is derived:

[0035]

[0036] Where M is the coefficient matrix, and

[0037] The stability of the discrete quasi-proportional resonance extended state observer (QR-ESO) depends on the eigenvalues ​​of the coefficient matrix M, which are the poles of the observer. If all eigenvalues ​​lie within the unit circle in the z-plane, the QR-ESO is stable. The characteristic polynomial of matrix M is:

[0038]

[0039] Where a1, a2, a3, a4 are the coefficients of the fourth-order characteristic polynomial;

[0040] According to the Jury stability criterion, discrete QR-ESO is stable under the following conditions:

[0041]

[0042] Substituting the selected observer parameters into the inequalities in the formula verifies that all stability conditions are satisfied, and the discrete quasi-proportional resonance extended state observer QR-ESO is stable.

[0043] Compared with existing technologies, the advantages of this invention are as follows: This invention employs an analytically precise discrete model in the αβ coordinate system, fully considering the rotor electrical angle variation within the sampling period, thus reducing the prediction error caused by the Euler approximation in the traditional dq coordinate system. Simultaneously, it uses QR-ESO for online estimation and compensation of parameter mismatch and external disturbances, effectively reducing the impact of stator resistance, inductance, and permanent magnet flux linkage variations on control performance. Furthermore, this invention can simultaneously compensate for DC and AC disturbances, improving the current tracking error and ripple increase under low carrier ratio conditions, thereby significantly improving the dynamic response, steady-state accuracy, and robustness of the permanent magnet synchronous motor predictive current control system. Compared with existing deadbeat predictive current control (DPCC) methods, this invention has better low carrier ratio adaptability, parameter robustness, and disturbance compensation capability. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0045] Figure 1 This is the overall control block diagram of the present invention.

[0046] Figure 2 The control block diagram for the quasi-proportional resonance extended state observer.

[0047] Figure 3 This is a waveform diagram of the direct and quadrature axis currents when the inductor is mismatched.

[0048] Figure 4 This is a waveform diagram of the direct and quadrature axis currents when there is resistor mismatch.

[0049] Figure 5 The waveforms of the direct and quadrature axis currents when the rotor flux linkage is mismatched are shown. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] like Figure 1 The method for deadbeat predictive current control based on a quasi-proportional resonant state observer, as shown, includes the following steps:

[0052] 1) The three-phase stator current, rotor electrical angle, and rotor electrical angular velocity of the permanent magnet synchronous motor (PMSM) are collected. The three-phase stator current is transformed to a two-phase stationary coordinate system to obtain the α-axis current and β-axis current. The voltage equation of the PMSM is transformed from the rotating dq coordinate system to the stationary αβ coordinate system to obtain a continuous-time current differential equation with the αβ-axis stator current as the state variable. In this model, the stator resistance voltage drop, voltage input term, and back electromotive force term generated by the permanent magnet flux linkage and rotor electrical angular velocity are explicitly retained. Subsequently, assuming that the stator voltage and electrical angular velocity are approximately constant within one sampling period, and considering the continuous change of the rotor electrical angle with time, the αβ-axis current differential equation is analytically integrated to obtain the accurate state transition relationship from the current sampling time to the next sampling time.

[0053] 2) In a stationary two-phase stationary coordinate system, an accurate discrete current prediction model is established based on the continuous-time voltage equation of the permanent magnet synchronous motor;

[0054] The accurate discrete current prediction model is established in a stationary αβ coordinate system, and its form is as follows:

[0055]

[0056]

[0057] In the formula, i α (k), u α (k) represent the α-axis current and α-axis voltage at time k, respectively. β (k), u β (k) represents the β-axis current and β-axis voltage at time k, A is the current state transition matrix, B is the voltage input matrix, G(k) is the precisely discrete term of the back electromotive force related to the permanent magnet flux linkage, rotor electric angular velocity, and rotor electric angle, and L s For stator inductance; R is the sampling period; s Stator resistance; ψ f For rotor flux linkage, ω e (k) represents the electrical angular velocity at time k;

[0058] The exact discrete current prediction model is rewritten as a voltage equation, and a one-step lead-delay compensation is used to offset the delay introduced by digital control, resulting in the following equation:

[0059]

[0060] Among them, u α (k+1) represents the predicted value of the α-axis voltage, u β (k+1) represents the predicted value of the β-axis voltage, i α (k+1) represents the predicted value of the α-axis current, i β (k+1) represents the predicted value of the β-axis current, i α (k+2) is the reference value for the α-axis current, i β (k+2) is the reference value of the β-axis current, and G[k+1] is the precise discrete term of the back electromotive force at time k+1 related to the permanent magnet flux linkage, rotor electric angular velocity and rotor electric angle.

[0061] Since the voltage equation model is derived from the αβ-axis current equation, no forward Euler approximation is introduced during discretization. Therefore, compared to the traditional dq-axis deadbeat predictive current control (DPCC), this model can more accurately describe the current evolution within a sampling period under low carrier ratio conditions. This significantly reduces the current prediction error.

[0062] 3) Construct a quasi-proportional resonance extended state observer QR-ESO (e.g.) Figure 2 The total disturbance caused by motor parameter mismatch, model uncertainty, sampling delay and low carrier ratio operation is observed online as an extended state to obtain the αβ axis disturbance estimate; one-step current prediction is performed based on the accurate discrete current prediction model, and delay compensation is performed in combination with the one-step delay characteristics of the digital control system;

[0063] The construction of a quasi-proportional resonance extended state observer (QR-ESO) specifically includes:

[0064]

[0065] Where x1 is the observed system output, and x2 is the total disturbance. 1 and 2 represents the observed values ​​of x1 and x2, respectively; β1 and β2 are the gains of the quasi-proportional resonant extended state observer QR-ESO, expressed as β1 = ξω0 and β2 = ω0. 2 ξ is the tuning coefficient that adjusts the dynamic response, ω0 is the bandwidth of the quasi-proportional resonant extended state observer QR-ESO; G(s) is the transfer function of the quasi-resonant controller.

[0066]

[0067] In the formula, k r For the resonant gain, ω r For the selected resonant frequency, ω c The cutoff frequency is given; the transfer function of the quasi-proportional resonant extended state observer QR-ESO is:

[0068] ;

[0069] The α-axis and β-axis observers of the quasi-proportional resonant extended state observer QR-ESO have the same structure. For simplicity, only the α-axis observer is given. The α-axis observer introduces the lumped disturbance as an extended state into the α-axis current equation, resulting in the α-axis permanent magnet synchronous motor model used for observer design:

[0070]

[0071] In the formula, i α F is the α-axis current value. α This represents the lumped disturbance value along the α-axis. and is i α and F α The estimated value; in this formula, the lumped disturbance includes both the DC component caused by parameter mismatch and the AC component caused by low carrier ratio operation. Therefore, the QR-ESO is constructed to simultaneously estimate these two types of disturbances. The above formula is further rewritten as:

[0072]

[0073] in, The DC component is caused by parameter mismatch. This refers to the AC component caused by low carrier ratio operation.

[0074] Discretizing the aforementioned observer model, the discrete quasi-proportional resonant extended state observer QR-ESO is expressed as follows:

[0075]

[0076] By adjusting the observer parameters, the total α-axis disturbance F is estimated. α (k+1), thus obtaining the α-axis current i α The current equation is rewritten in voltage form, and a one-step delay compensation e is introduced. α Then, the equation at time k+1 is expressed as:

[0077]

[0078] The quasi-proportional resonant extended state observer QR-ESO satisfies the Jury stability criterion, and its stability proof is as follows: Using the discrete formula of the QR-ESO, where the α-axis current i is chosen... α Total disturbance F along the α axis α The quasi-resonant controller's state variables x1 and x2 are used as state variables:

[0079]

[0080]

[0081] The stability of the discrete quasi-proportional resonant extended state observer (QR-ESO) depends on the eigenvalues ​​of the coefficient matrix M, which are the poles of the observer. If all eigenvalues ​​lie within the unit circle in the z-plane, the QR-ESO is stable; the characteristic polynomial of matrix M is:

[0082]

[0083] Where a1, a2, a3, a4 are the coefficients of the fourth-order characteristic polynomial;

[0084] The stability of the discrete quasi-proportional resonant extended state observer QR-ESO, according to the Jury stability criterion, satisfies the following condition:

[0085]

[0086] Substituting the selected observer parameters into the inequalities in the equation verifies that all stability conditions are satisfied. Therefore, the proposed discrete proportional resonant state observer is stable.

[0087] 4) The αβ axis reference voltage, accurately predicted by the voltage vector, is input to the space vector pulse width modulated (PWM). The output PWM wave is processed by the inverter to obtain the drive voltage, which drives the permanent magnet synchronous motor. The actual speed of the permanent magnet synchronous motor during operation is detected by a position sensor. The difference between the actual speed and the given speed is then used by PI control to output the reference speed i. q Value, and reference i d The values ​​are transformed by coordinates to obtain the αβ axis reference current for voltage vector prediction, forming a speed closed-loop control system;

[0088] The αβ axis reference voltage u is obtained by accurately predicting the voltage vector. αβ k+1 The input is SVPWM (Space Vector Pulse Width Modulation), and the output PWM wave is sent to the inverter. After processing by the inverter, the drive voltage is obtained, which drives the permanent magnet synchronous motor to operate.

[0089] The three-phase current i is sampled from the inverter. abc The actual current i along the αβ axis is obtained through a static transformation matrix. αβ The static coordinate transformation matrix is ​​as follows:

[0090]

[0091] In vector control, it is necessary to perform coordinate transformation on voltage or current in different coordinate systems, so determining the transformation matrix is ​​very important.

[0092] The actual rotational speed of the permanent magnet synchronous motor is detected by a position sensor. The difference between the actual speed and the given speed is then used to output a reference speed via PI control. q Value, and reference i d The values ​​are transformed to obtain the αβ-axis reference current i for voltage vector prediction. αβ k+2 The coordinate transformation matrix is ​​as follows:

[0093]

[0094] The obtained reference current i αβ k+2 The input voltage vector accurate prediction module forms a speed closed-loop control system.

[0095] Figure 3 , Figure 4 and Figure 5 The figures show the direct and quadrature axis current waveforms of the quasi-proportional resonant extended state observer under stator inductance mismatch, stator resistance mismatch, and rotor flux mismatch conditions, respectively. As can be seen from the figures, the accurate prediction model based on the quasi-proportional resonant extended state observer can accurately compensate for disturbances while effectively improving the problems of current tracking error and increased ripple under low carrier ratio conditions, thereby significantly improving the dynamic response, steady-state accuracy, and robustness of the permanent magnet synchronous motor predictive current control system.

[0096] The above description of the embodiments is only for the purpose of helping to understand the method and core ideas of the present invention. It should be noted that those skilled in the art can make several improvements and modifications to the present invention without departing from the principles of the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Claims

1. A deadbeat predictive current control method based on a quasi-proportional resonant state observer, characterized in that, Includes the following steps: 1) Collect the three-phase stator current, rotor electrical angle and rotor electrical angular velocity of the permanent magnet synchronous motor, and transform the three-phase stator current to a two-phase stationary coordinate system to obtain the α-axis current and β-axis current; 2) In a stationary two-phase stationary coordinate system, an accurate discrete current prediction model is established based on the continuous-time voltage equation of the permanent magnet synchronous motor; 3) Construct a quasi-proportional resonance extended state observer (QR-ESO) to observe the total disturbance caused by motor parameter mismatch, model uncertainty, sampling delay, and low carrier ratio operation as an extended state online, and obtain the αβ axis disturbance estimate; perform one-step current prediction based on the accurate discrete current prediction model, and perform delay compensation in combination with the one-step delay characteristics of the digital control system; 4) The αβ axis reference voltage, accurately predicted by the voltage vector, is input to the space vector pulse width modulated (PWM). The output PWM wave is processed by the inverter to obtain the drive voltage, which drives the permanent magnet synchronous motor. The actual speed of the permanent magnet synchronous motor during operation is detected by a position sensor. The difference between the actual speed and the given speed is then used by PI control to output the reference speed i. q Value, and reference i d The values ​​are transformed by coordinates to obtain the αβ axis reference current for voltage vector prediction, forming a speed closed-loop control system.

2. The deadbeat predictive current control method based on a quasi-proportional resonant state observer according to claim 1, characterized in that, Step 2) describes establishing an accurate discrete current prediction model based on the continuous-time voltage equation of the permanent magnet synchronous motor, which specifically includes: The precise discrete current prediction model is established in a stationary αβ coordinate system, and its form is as follows: ； ； In the formula, i α (k), u α (k) represent the α-axis current and α-axis voltage at time k, respectively. β (k), u β (k) represents the β-axis current and β-axis voltage at time k, A is the current state transition matrix, B is the voltage input matrix, G(k) is the precisely discrete term of the back electromotive force related to the permanent magnet flux linkage, rotor electric angular velocity, and rotor electric angle, and L s For stator inductance; R is the sampling period; s Stator resistance; ψ f For rotor flux linkage, ω e (k) represents the electrical angular velocity at time k; The exact discrete current prediction model is rewritten as a voltage equation, and a one-step lead-delay compensation is used to offset the delay introduced by digital control, resulting in the following equation: ； Among them, u α (k+1) represents the predicted value of the α-axis voltage, u β (k+1) represents the predicted value of the β-axis voltage, i α (k+1) represents the predicted value of the α-axis current, i β (k+1) represents the predicted value of the β-axis current, i α (k+2) is the reference value for the α-axis current, i β (k+2) is the reference value of the β-axis current, and G[k+1] is the precise discrete term of the back electromotive force at time k+1 related to the permanent magnet flux linkage, rotor electric angular velocity and rotor electric angle.

3. The deadbeat predictive current control method based on a quasi-proportional resonant state observer according to claim 1, characterized in that, Step 3) describes the construction of the quasi-proportional resonance extended state observer QR-ESO, which specifically includes: ； Where x1 is the observed system output, and x2 is the total disturbance. and x1 and x2 are the observed values, respectively; β1 and β2 are the gains of the quasi-proportional resonant extended state observer QR-ESO, expressed as β1 = ξω0 and β2 = ω0. 2 ξ is the tuning coefficient that adjusts the dynamic response, ω0 is the bandwidth of the quasi-proportional resonant extended state observer QR-ESO; G(s) is the transfer function of the quasi-resonant controller. ； In the formula, k r For the resonant gain, ω r For the selected resonant frequency, ω c The cutoff frequency is given; the transfer function of the quasi-proportional resonant extended state observer QR-ESO is: 。 4. The deadbeat predictive current control method based on a quasi-proportional resonant state observer according to claim 3, characterized in that, The α-axis and β-axis observers of the quasi-proportional resonant extended state observer QR-ESO have the same structure. The α-axis observer introduces the lumped disturbance as an extended state into the α-axis current equation, resulting in an α-axis permanent magnet synchronous motor model for observer design: ； In the formula, i α F is the α-axis current value. α This represents the lumped disturbance value along the α-axis. and is i α and F α The estimated value; The above formula can be further rewritten as: ； in, This refers to the DC component caused by parameter mismatch. The AC component caused by low carrier ratio operation; Discretizing the aforementioned observer model, the discrete quasi-proportional resonant extended state observer QR-ESO is expressed as follows: ； By adjusting the observer parameters, the total α-axis disturbance F is estimated. α (k+1), thus obtaining the α-axis current i α The current equation is rewritten in voltage form, and a one-step delay compensation e is introduced. α Then, the equation at time k+1 is expressed as: 。 5. The deadbeat predictive current control method based on a quasi-proportional resonant state observer according to claim 4, characterized in that, The stability of the discrete quasi-proportional resonance extended state observer QR-ESO is based on the Jury stability criterion, selecting the α-axis current i. α Total disturbance F along the α axis α Using the quasi-resonant controller state variables x1 and x2 as state variables, the discrete formula for QR-ESO is derived: ； Where M is the coefficient matrix, and ; The stability of the discrete quasi-proportional resonance extended state observer (QR-ESO) depends on the eigenvalues ​​of the coefficient matrix M, which are the poles of the observer. If all eigenvalues ​​lie within the unit circle in the z-plane, the QR-ESO is stable. The characteristic polynomial of matrix M is: ； Where a1, a2, a3, a4 are the coefficients of the fourth-order characteristic polynomial; ； According to the Jury stability criterion, discrete QR-ESO is stable under the following conditions: ； Substituting the selected observer parameters into the inequalities in the formula verifies that all stability conditions are satisfied, and the discrete quasi-proportional resonance extended state observer QR-ESO is stable.

Images