Dead-beat predictive current control method based on expanded notch state observer

Abstract

The invention discloses a dead-beat predictive current control method based on an expanded notch state observer, and relates to the field of permanent magnet synchronous motor current control, and the method comprises the steps: designing a cascaded expanded state observer combined with a notch filter; and designing a dead-beat predictive current control method of the permanent magnet synchronous motor based on the cascaded extended state observer combined with the notch filter, and completing dead-beat predictive current control based on the extended notch state observer. According to the invention, the problems of periodic disturbance amplification and complex stability judgment near the resonant frequency in the prior art are solved.

Description

Technical Field

[0001] The invention relates to the field of permanent magnet synchronous motor current control, and in particular to a deadbeat predictive current control method based on an extended notch state observer. Background Art

[0002] Permanent magnet synchronous motor deadbeat predictive current control has been widely studied and applied in the field of permanent magnet synchronous motor control due to its strong dynamic performance. However, deadbeat predictive current control is a typical model-based control method, and its prediction accuracy depends on the accuracy of the motor model. The inaccuracy of the motor model will affect the current prediction accuracy, thereby affecting the dynamic performance and steady-state performance of the system. In actual motor applications, factors such as temperature changes and flux saturation will affect the electrical parameters of the motor. Inverter nonlinearity and current sampling errors introduce complex disturbances to the motor model. How to deal with the prediction error caused by the inaccuracy of the motor model has become a current research hotspot.

[0003] The method based on the disturbance observer compensation is based on the idea of ​​disturbance estimation-compensation. The unknown disturbance in the motor model is observed and compensated into the prediction model to reduce the impact of the disturbance on the system control performance. The prior art proposes an adaptive integral extended state observer to compensate for the disturbance caused by parameter mismatch and unmodeled items, and establishes a bandwidth adaptive law to dynamically adjust the bandwidth of the observer. When the observation error is large, the rapid convergence of the observer is achieved. When the observer error is small, the smaller bandwidth of the observer avoids the amplification of high-frequency noise caused by the excessive bandwidth of the observer. However, in steady state, affected by the small bandwidth, the observer's ability to suppress periodic disturbances is still limited. The prior art also proposes to enhance the system's ability to suppress low-frequency disturbances by cascading extended state observers, but the ability to suppress periodic disturbances is still limited. The prior art also proposes to embed a resonant controller into the design of an extended state observer to achieve accurate estimation of periodic disturbances of a specific frequency. However, periodic disturbances near the resonant frequency will be amplified, which deteriorates the control performance of the system to a certain extent. And when suppressing periodic disturbances of multiple different frequencies, the stability judgment of this method is complex.

[0004] However, the existing technologies for the extended state observer combined with different controllers all have common problems such as amplification of periodic disturbances near the resonant frequency and complex stability judgment. This deteriorates the control performance of the system to a certain extent, especially when the frequency of the periodic disturbance to be suppressed is high, the existing solutions inevitably amplify the high-frequency noise of the system, which in turn deteriorates the control performance of the system. Summary of the invention

[0005] In view of the above-mentioned deficiencies in the prior art, the deadbeat predictive current control method based on an extended notch state observer provided by the present invention solves the problems of periodic disturbance amplification near the resonant frequency and complex stability judgment existing in the prior art.

[0006] In order to achieve the above-mentioned invention object, the technical solution adopted by the present invention is: a deadbeat predictive current control method based on an extended notch state observer, comprising the following steps: S1: Design a cascade extended state observer combined with a notch filter; S2: Based on the cascade extended state observer combined with the notch filter, a beat-free predictive current control method for the permanent magnet synchronous motor is designed to complete the beat-free predictive current control based on the extended notch state observer.

[0007] Furthermore, the cascade extended state observer combined with the notch filter in S1 includes q Extended state observer for shaft DC disturbance, q Shaft harmonic expansion state observer, d Extended state observer for shaft DC disturbance and d Shaft harmonic expansion state observer.

[0008] Furthermore, the q The extended state observer of shaft DC disturbance is:

[0009] in, for q Estimation error of the extended state observer for shaft DC disturbance, Permanent magnet synchronous motor q Shaft current, for The observed value of for q The observed value of the shaft DC disturbance, is the reciprocal of the inductance of the permanent magnet synchronous motor, Permanent magnet synchronous motor q Shaft voltage, for q The known axis quantity, and is the DC disturbance extended state observer gain, represents the first-order derivative.

[0010] Furthermore, the q The shaft harmonic expansion state observer is:

[0011] in, for q Axisn A harmonic extended state observer observes the error, for The observed value of for q Axis ( ) harmonic disturbance observation value of the harmonic extended state observer, when =0, for q The DC disturbance observation value of the axis DC disturbance observer, is the number of harmonic extended state observers, is the harmonic extended state observer gain, for q The frequency observed on the axis is The harmonic disturbance of and is an auxiliary variable, and is an auxiliary variable, and is an auxiliary variable;

[0012] in, is the notch filter notch bandwidth, is the notch depth of the notch filter.

[0013] Furthermore, the d The extended state observer of shaft DC disturbance is:

[0014] in, for d Estimation error of the extended state observer for shaft DC disturbance, Permanent magnet synchronous motor d Shaft current, for The observed value of for d The observed value of the shaft DC disturbance, Permanent magnet synchronous motor d Shaft voltage, for d Axis known quantity.

[0015] Furthermore, the d The shaft harmonic expansion state observer is:

[0016] in, for d Axis nA harmonic extended state observer observes the error, for The observed value of for d Axis ( ) harmonic disturbance observation value of the harmonic extended state observer, when =0, for d The DC disturbance observation value of the axis DC disturbance observer, is the number of harmonic extended state observers, for d The frequency observed on the axis is The harmonic disturbance of and is an auxiliary variable, and is an auxiliary variable, and is an auxiliary variable; .

[0017] Furthermore, the permanent magnet synchronous motor zero-beat prediction current control method is designed in S2, and the formula is:

[0018] in, is the current, is the current control moment, For the next control moment, is the state matrix, is the input parameter, is the additional input, is the input voltage, is the disturbance estimate;

[0019]

[0020]

[0021]

[0022]

[0023] in, To control the cycle, is the inductor, and Permanent Magnet Synchronous Motor d and q Shaft current, and Permanent Magnet Synchronous Motor d and q Shaft voltage, and They are d and q Axis disturbance estimate, is the resistance, is the electrical angular velocity, is the magnetic chain, the superscript Transpose the matrix; think , calculate the reference voltage:

[0024]

[0025]

[0026] in, is the reference input voltage, is the reference current, and They are d and q Axis reference input voltage, and They are d and q Axis reference input current; The predicted current is made equal to the reference input value according to the reference voltage, thereby realizing predicted current control.

[0027] The beneficial effects of the present invention are: (1) The present invention embeds a notch filter into the extended state observer to achieve accurate estimation of specific frequencies and DC interference.

[0028] (2) The cascade design of the present invention separates the design of the DC disturbance extended state observer and the harmonic extended state observer, realizes the decoupling of the estimation of the DC disturbance and the periodic disturbance, and has a simple structure.

[0029] (3) The present invention solves the problem of interference amplification near the resonant frequency in the traditional method and prevents the influence of interference of other frequencies on the control performance, especially high-frequency noise. At the same time, the present invention has the advantages of simple stability analysis and parameter adjustment. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of the deadbeat predictive current control method based on the expanded notch state observer of the present invention.

[0031] Figure 2It is an overall control block diagram of the deadbeat predictive current control method based on the expanded notch state observer of the present invention.

[0032] Figure 3 is the transfer function In different Bode diagram below.

[0033] Figure 4 is the transfer function In different Bode diagram below.

[0034] Figure 5 is the transfer function In different Bode diagram below. DETAILED DESCRIPTION

[0035] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0036] like Figure 1 As shown, a deadbeat predictive current control method based on an extended notch state observer comprises the following steps: S1: Design a cascade extended state observer combined with a notch filter; S2: Based on the cascade extended state observer combined with the notch filter, a beat-free predictive current control method for the permanent magnet synchronous motor is designed to complete the beat-free predictive current control based on the extended notch state observer.

[0037] In one embodiment of the present invention, a permanent magnet synchronous motor considering unknown disturbances dq The shaft current equation is:

[0038] in, , They are dq Axis lumped disturbance; For simplicity, rewrite the above formula as follows:

[0039]

[0040] The cascade extended state observer combined with the notch filter proposed in the present invention can be designed in two steps. The first step is to design a DC disturbance extended state observer, and the second step is to design a harmonic extended state observer.

[0041] The cascade extended state observer combined with the notch filter in S1 includes q Extended state observer for shaft DC disturbance, q Shaft harmonic expansion state observer, dExtended state observer for shaft DC disturbance and d Shaft harmonic expansion state observer.

[0042] Said q The extended state observer of shaft DC disturbance is:

[0043] in, for q Estimation error of the extended state observer for shaft DC disturbance, Permanent magnet synchronous motor q Shaft current, for The observed value of for q The observed value of the shaft DC disturbance, is the reciprocal of the inductance of the permanent magnet synchronous motor, Permanent magnet synchronous motor q Shaft voltage, for q The known axis quantity, and is the DC disturbance extended state observer gain, represents the first-order derivative.

[0044] Said q The shaft harmonic expansion state observer is:

[0045] in, for q Axis n A harmonic extended state observer observes the error, for The observed value of for q Axis ( ) harmonic disturbance observation value of the harmonic extended state observer, when =0, for q The DC disturbance observation value of the axis DC disturbance observer, is the number of harmonic extended state observers, is the harmonic extended state observer gain, for q The frequency observed on the axis is The harmonic disturbance of and is an auxiliary variable, and is an auxiliary variable, and is an auxiliary variable; The entire observer is composed of a DC disturbance observer and multiple harmonic disturbance observers. Numbered from 0 to n, =0 refers to the observed value of this DC disturbance observer. =1, which is the The observed values ​​of the harmonic disturbance observer;

[0046] in, is the notch filter notch bandwidth, is the notch depth of the notch filter.

[0047] Said d The extended state observer of shaft DC disturbance is:

[0048] in, for d Estimation error of the extended state observer for shaft DC disturbance, Permanent magnet synchronous motor d Shaft current, for The observed value of for d The observed value of the shaft DC disturbance, Permanent magnet synchronous motor d Shaft voltage, for d Axis known quantity.

[0049] Said d The shaft harmonic expansion state observer is:

[0050] in, for d Axis n A harmonic extended state observer observes the error, for The observed value of for d Axis ( ) harmonic disturbance observation value of the harmonic extended state observer, when =0, for d The DC disturbance observation value of the axis DC disturbance observer, is the number of harmonic extended state observers, for d The frequency observed on the axis is The harmonic disturbance of and is an auxiliary variable, and is an auxiliary variable, and is an auxiliary variable; .

[0051] The forward Euler formula is used to make one-step predictions to compensate for the one-step delay caused by the microprocessor.

[0052] The permanent magnet synchronous motor zero-beat prediction current control method is designed in S2, and the formula is:

[0053] in, is the current, is the current control moment, For the next control moment, is the state matrix, is the input parameter, is the additional input, is the input voltage, is the disturbance estimate;

[0054]

[0055]

[0056]

[0057]

[0058]

[0059]

[0060] in, To control the cycle, is the inductor, and Permanent Magnet Synchronous Motor d and q Shaft current, and Permanent Magnet Synchronous Motor d and q Shaft voltage, and They are d and q Axis disturbance estimate, is the resistance, is the electrical angular velocity, is the magnetic chain, the superscript Transpose the matrix; think , calculate the reference voltage:

[0061]

[0062]

[0063] in, is the reference input voltage, is the reference current, and They are d and q Axis reference input voltage, and They are d and q Axis reference input current; The predicted current is made equal to the reference input value according to the reference voltage, thereby realizing predicted current control.

[0064] The control block diagram of the permanent magnet synchronous motor of the present invention is as follows: Figure 2 As shown, for dq The present invention provides a cascade extended state observer combined with a notch filter to observe the DC, low-frequency disturbance components and harmonic components in the unknown disturbance. The cascade ESO combined with the notch filter includes a DC disturbance ESO and a harmonic ESO, which form a total observer in a cascade form. Among them, the DC disturbance ESO is used to observe the DC and low-frequency disturbance components in the disturbance, and the harmonic ESO is used to observe the specific subharmonic components in the disturbance. The total disturbance observed by the observer is compensated into the deadbeat predictive current control (DPCC) algorithm to achieve the correction of the output voltage.

[0065] In one embodiment of the present invention, q axis as an example, when it is necessary to observe DC disturbances and the frequency is When the harmonic is , the disturbance observation error can be obtained according to its DC disturbance extended state observer and harmonic extended state observer With actual q Axis disturbance The transfer function between:

[0066] in, is a complex variable.

[0067] Plot the above equation about different , and The Bode diagram of Figure 3-5 As shown in FIG. 1 , it can be found that the method proposed in the present invention has no amplification effect on the disturbance near the notch frequency, which solves the problem of the traditional method amplifying the disturbance near the resonant frequency, and can be and The freely adjustable notch bandwidth and notch depth enable more accurate observation of harmonic disturbances, and the large bandwidth setting makes harmonic extraction less sensitive to harmonic frequencies.

[0068] When there are multiple harmonics to be observed, the disturbance observation error transfer function becomes:

[0069] It can be seen that the transfer function is realized by continuous multiplication, which is simple in form and can be adjusted using the bandwidth method. and ,Right now , , is the observer bandwidth. From the form of the transfer function, we only need to ensure , and This ensures system stability.

[0070] The present invention provides an improved deadbeat current predictive control technology based on a cascaded extended state observer combined with a notch filter, which can achieve accurate DC disturbance and multiple specific frequency harmonic disturbance suppression. Through the cascade structure, the estimation of DC disturbance and harmonic disturbance is decoupled, avoiding the amplification of disturbance near the notch frequency. In addition, the method proposed by the present invention simplifies the transfer function of the system, reduces the complexity of system stability analysis, and greatly simplifies parameter tuning.

[0071] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the invention.

Claims

1. A deadbeat predictive current control method based on an extended notch state observer, characterized in that: The following steps are involved: S1: Design a cascade extended state observer combined with a notch filter; S2: Based on the cascade extended state observer combined with the notch filter, a beat-free predictive current control method for the permanent magnet synchronous motor is designed to complete the beat-free predictive current control based on the extended notch state observer.

2. The deadbeat predictive current control method based on an extended notch state observer according to claim 1, characterized in that: The cascade extended state observer combined with the notch filter in S1 includes q Extended state observer for shaft DC disturbance, q Shaft harmonic expansion state observer, d Extended state observer for shaft DC disturbance and d Shaft harmonic expansion state observer.

3. The deadbeat predictive current control method based on an extended notch state observer according to claim 2, characterized in that: Said q The extended state observer of shaft DC disturbance is: in, for q Estimation error of the extended state observer for shaft DC disturbance, Permanent magnet synchronous motor q Shaft current, for The observed value of for q The observed value of the shaft DC disturbance, is the reciprocal of the inductance of the permanent magnet synchronous motor, Permanent magnet synchronous motor q Shaft voltage, for q The known axis quantity, and is the DC disturbance extended state observer gain, represents the first-order derivative.

4. The deadbeat predictive current control method based on an extended notch state observer according to claim 3, characterized in that: Said q The shaft harmonic expansion state observer is: in, for q Axis n A harmonic extended state observer observes the error, for The observed value of for q Axis ( ) harmonic disturbance observation value of the harmonic extended state observer, when =0, for q The DC disturbance observation value of the axis DC disturbance observer, is the number of harmonic extended state observers, is the harmonic extended state observer gain, for q The frequency observed on the axis is The harmonic disturbance of and is an auxiliary variable, and is an auxiliary variable, and is an auxiliary variable; in, is the notch filter notch bandwidth, is the notch depth of the notch filter.

5. The deadbeat predictive current control method based on an extended notch state observer according to claim 4, characterized in that: Said d The extended state observer of shaft DC disturbance is: in, for d Estimation error of the extended state observer for shaft DC disturbance, Permanent magnet synchronous motor d Shaft current, for The observed value of for d The observed value of the shaft DC disturbance, Permanent magnet synchronous motor d Shaft voltage, for d Axis known quantity.

6. The deadbeat predictive current control method based on an extended notch state observer according to claim 5, characterized in that: Said d The shaft harmonic expansion state observer is: in, for d Axis n A harmonic extended state observer observes the error, for The observed value of for d Axis ( ) harmonic disturbance observation value of the harmonic extended state observer, when =0, for d The DC disturbance observation value of the axis DC disturbance observer, is the number of harmonic extended state observers, for d The frequency observed on the axis is The harmonic disturbance of and is an auxiliary variable, and is an auxiliary variable, and is an auxiliary variable; 。 7. The deadbeat predictive current control method based on an extended notch state observer according to claim 6, characterized in that: The permanent magnet synchronous motor zero-beat prediction current control method is designed in S2, and the formula is: in, is the current, is the current control moment, For the next control moment, is the state matrix, is the input parameter, is the additional input, is the input voltage, is the disturbance estimate; in, To control the cycle, is the inductor, and Permanent Magnet Synchronous Motor d and q Shaft current, and Permanent Magnet Synchronous Motor d and q Shaft voltage, and They are d and q Axis disturbance estimate, is the resistance, is the electrical angular velocity, is the magnetic chain, the superscript Transpose the matrix; think , calculate the reference voltage: in, is the reference input voltage, is the reference current, and They are d and q Axis reference input voltage, and They are d and q Axis reference input current; The predicted current is made equal to the reference input value according to the reference voltage, thereby realizing predicted current control.

Images