A deadbeat current prediction control method and system based on integral compensation

Abstract

The present invention discloses a deadbeat current prediction control method and system based on integral compensation, which belongs to the technical field of permanent magnet synchronous motor control. In view of the error in inductance parameters, the d-axis current error integral is introduced into the d-axis control voltage result to solve the current static error problem caused by the inductance parameter error; in view of the error in permanent magnet flux linkage parameters, according to i q The current error is proportional to the flux error. q The error in the flux linkage is used to adjust the value of the flux linkage so that the flux linkage parameters converge to the real value in the motor, eliminating the current static error problem caused by the flux linkage error. The present invention can retain the advantages of the high bandwidth and simple calculation of the zero-beat current prediction control algorithm when there are errors between the estimated parameters of the system model motor inductance and the permanent magnet flux linkage and the actual parameters, while solving the problem of the dq-axis current static error caused by the motor parameter error.

Description

Technical Field

[0001] The present invention belongs to the technical field of permanent magnet synchronous motor control, and more specifically, relates to a deadbeat current prediction control method and system based on integral compensation. Background Art

[0002] With the rapid development of today's industrial field, AC servo systems based on permanent magnet synchronous motors have occupied a vital position in the fields of electric vehicles, robot control technology and CNC machine tools. At the same time, with the continuous development of these fields, the industry has also put forward higher requirements on the performance of permanent magnet synchronous motor AC servo systems, not only hoping that they have faster response speed and higher working accuracy, but also hoping to further improve their working reliability. Therefore, for AC servo systems, the selection of control strategies and performance improvement are particularly important.

[0003] There are three main control closed loops in the permanent magnet synchronous motor drive based on vector control, namely the position loop, speed loop and current loop. As the innermost loop of the system, the speed of the dynamic response performance of the current loop has the greatest impact on the overall performance of the system, and its bandwidth is an important factor affecting the dynamic performance of the servo system.

[0004] The most commonly used current loop control scheme in the field of motor control is the PI (proportional integral) controller. However, due to the characteristics of the PI algorithm, the current loop has problems such as slow response, large overshoot, and poor control performance. Deadbeat current predictive control has become a hot topic in servo motor control algorithm research in recent years because of its advantages such as fast dynamic response, simple and intuitive algorithm, no need to set parameters, small amount of calculation, and fixed switching frequency.

[0005] Deadbeat current predictive control is a control algorithm based on the mathematical model of the motor. In practical applications, the quality of its control performance depends largely on the accuracy of the motor parameters in the mathematical model. The mismatch of the motor parameters will cause the dq axis current to be unable to accurately follow the command value, thus affecting the dynamic and static performance of the system. To address this problem, the current mainstream solution is to add motor parameter identification or observers to the system to correct the parameters in the model in real time or enhance the ability of the prediction model to resist parameter disturbances. However, both of these methods and algorithms will greatly increase the complexity of the system algorithm, limiting its practical application. Summary of the invention

[0006] In view of the above defects or improvement needs of the prior art, the present invention provides a beat-free current prediction control method and system based on integral compensation, which aims to solve the technical problem of dq-axis current static error caused by motor parameter error in beat-free current prediction control.

[0007] To achieve the above object, according to one aspect of the present invention, a deadbeat current prediction control method based on integral compensation is provided, comprising:

[0008] S1. According to the actual value of the motor q-axis current i q With the instruction value i q * The error between them is used to perform integral compensation on the permanent magnet flux parameters of the motor;

[0009] S2. Take the motor d-axis current command value, q-axis current command value and the motor permanent magnet flux parameters after integral compensation as input, perform deadbeat current predictive control on the motor, and obtain the motor d and q axis control voltage u d 、u q ;

[0010] S3. According to the actual value of motor d-axis current i d With the instruction value i d * The error between them controls the motor d-axis voltage u d Make points compensation;

[0011] S4. Use the motor q axis to control the voltage u q And the motor d-axis control voltage u after integral compensation d The motor is pulse-width modulated using space vector technology.

[0012] Furthermore, the actual value of the motor q-axis current i q With the instruction value i q * The error between and the permanent magnet flux parameter error satisfies the following relationship:

[0013]

[0014] L0 is the actual value of the inductance parameter, T is the motor torque, ω e (k) is the △ψ of the motor at the kth moment f is the permanent magnet flux parameter error.

[0015] Furthermore, the following formula is used to perform integral compensation on the permanent magnet flux parameters of the motor:

[0016]

[0017] ψ f com is the permanent magnet flux of the motor after integral compensation, ψ f is the permanent magnet flux of the motor, The q-axis current command value of the motor at the kth moment, i q (k) is the actual value of the motor's q-axis current at the kth moment, K iψis the flux integration coefficient.

[0018] Furthermore, the following formula is used to control the motor d-axis voltage u d To make points compensation:

[0019]

[0020] u d (k) com is the d-axis control voltage of the motor in the kth control cycle after integral compensation, u d (k) is the d-axis control voltage of the motor in the kth control cycle, The d-axis current command value of the motor at the kth moment, i d (k) is the actual value of the motor's d-axis current at the kth moment, K id is the d-axis voltage integral coefficient.

[0021] According to another aspect of the present invention, there is provided a deadbeat current prediction control system based on integral compensation, comprising: a flux compensation module, a deadbeat current prediction controller, a d-axis control voltage compensation module and a SVPWM modulation module;

[0022] The flux compensation module is used to compensate the actual value of the motor q-axis current i q With the instruction value i q * The error between them is used to perform integral compensation on the permanent magnet flux parameters of the motor;

[0023] The deadbeat current prediction controller is used to perform deadbeat current prediction control on the motor with the motor d-axis current command value, q-axis current command value and the motor permanent magnet flux linkage parameter after integral compensation as input to obtain the motor d-axis and q-axis control voltage u d 、u q ;

[0024] The d-axis control voltage compensation module is used to adjust the actual value of the motor d-axis current i d With the instruction value i d * The error between them controls the motor d-axis voltage u d Make points compensation;

[0025] SVPWM modulation module, used to control the voltage u using the motor q axis q And the motor d-axis control voltage u after integral compensation d The motor is pulse-width modulated using space vector technology.

[0026] Furthermore, the flux compensation module performs integral compensation on the permanent magnet flux parameters of the motor through the following formula:

[0027]

[0028] ψ f com is the permanent magnet flux of the motor after integral compensation, ψ f is the permanent magnet flux of the motor, The q-axis current command value of the motor at the kth moment, i q (k) is the actual value of the motor's q-axis current at the kth moment, K iψ is the flux integration coefficient.

[0029] Furthermore, the d-axis control voltage compensation module calculates the motor d-axis control voltage u by the following formula: d To make points compensation:

[0030]

[0031] u d (k) com is the d-axis control voltage of the motor in the kth control cycle after integral compensation, u d (k) is the d-axis control voltage of the motor in the kth control cycle, The d-axis current command value of the motor at the kth moment, i d (k) is the actual value of the motor's d-axis current at the kth moment, K id is the d-axis voltage integral coefficient

[0032] In general, compared with the prior art, the above technical solution conceived by the present invention can achieve the following beneficial effects.

[0033] The present invention introduces the d-axis current error integral into the d-axis control voltage result to solve the current static error problem caused by the inductance parameter error; q The current error is proportional to the flux error. q The error in the flux is used to adjust the value of the flux linkage, so that the flux linkage parameters converge to the real value in the motor, eliminating the current static error problem caused by the flux linkage error. The present invention can retain the advantages of high bandwidth and simple calculation of the zero-beat current prediction control algorithm when there are errors between the estimated parameters of the system model motor inductance and the permanent magnet flux linkage and the actual parameters, while solving the problem of dq-axis current static error caused by the motor parameter error. Compared with the traditional PI-based current loop system, the present invention has better dynamic response performance, simple calculation, and is suitable for engineering application. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a control flow chart of the method of the present invention;

[0035] Figure 2(a) and (b) are the given command current i of the traditional PI current loop and the deadbeat current prediction control current loop with accurate motor parameters. q * Simulation results when the current increases from 1A to 2A at 0.1s and then decreases from 2A to 0.5A at 0.3s;

[0036] Figure 3 (a) and (b) are the simulation results before and after adding integral compensation after the deadbeat current prediction control algorithm introduces inductance parameter error;

[0037] Figure 4 (a) and (b) are the simulation results before and after adding integral compensation after the permanent magnet flux parameter error is introduced by the deadbeat current predictive control algorithm. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0039] Deadbeat current prediction control is a control algorithm based on the mathematical model of the motor. The quality of its control performance depends largely on the accuracy of the motor parameters in the mathematical model. The mismatch of the motor parameters will cause the dq axis current to be unable to accurately follow the command value, thereby affecting the dynamic and steady-state performance of the system. In order to achieve the purpose of ensuring that the deadbeat current prediction controller can still work normally when there are errors in the motor parameters, the present invention proposes an integral compensation method based on the derivation of the deadbeat current prediction model, and adds integral compensation links in the system for the errors of the motor inductance and permanent magnet flux parameters, thereby eliminating the current static error problem caused by parameter mismatch to the current loop.

[0040] refer to Figure 1 The present invention provides a deadbeat current prediction control method based on integral compensation, comprising the following steps:

[0041] S1. According to the actual value of the motor q-axis current i q With the instruction value i q * The error between them is used to perform integral compensation on the permanent magnet flux parameters of the motor;

[0042] S2. Take the motor d-axis current command value, q-axis current command value and the motor permanent magnet flux parameters after integral compensation as input, perform deadbeat current predictive control on the motor, and obtain the motor d and q axis control voltage ud 、u q ;

[0043] S3. According to the actual value of motor d-axis current i d With the instruction value i d * The error between them controls the motor d-axis voltage u d Make points compensation;

[0044] S4. Use the motor q axis to control the voltage u q And the motor d-axis control voltage u after integral compensation d The motor is pulse-width modulated using space vector technology.

[0045] The following formula is used to perform integral compensation on the permanent magnet flux parameters of the motor:

[0046]

[0047] ψ f com is the permanent magnet flux of the motor after integral compensation, ψ f is the permanent magnet flux of the motor, The q-axis current command value of the motor at the kth moment, i q (k) is the actual value of the motor's q-axis current at the kth moment, K iψ is the flux linkage integral coefficient.

[0048] Use the following formula to control the motor d-axis voltage u d To make points compensation:

[0049]

[0050] u d (k) com is the d-axis control voltage of the motor in the kth control cycle after integral compensation, u d (k) is the d-axis control voltage of the motor in the kth control cycle, The d-axis current command value of the motor at the kth moment, i d (k) is the actual value of the motor's d-axis current at the kth moment, K id is the d-axis voltage integral coefficient.

[0051] i d 、i q The acquisition process includes:

[0052] The three-phase stator current of the permanent magnet synchronous motor is sampled to obtain i a 、i b 、i c , the three-phase current is transformed by Clark to obtain the current i in the two-phase stationary α-β coordinate systemα 、i β ; The formula for Clark transformation is:

[0053]

[0054] Combined with the rotor angle θ returned by the permanent magnet synchronous motor position sensor, the current i in the two-phase stationary α-β coordinate system is obtained α 、i β Perform Park transformation to obtain i in the two-phase rotating dq coordinate system d 、i q ; The formula for Park transformation is:

[0055]

[0056] Step S4 specifically includes:

[0057] will u d 、u q Perform Park inverse transformation to obtain the current u in the two-phase stationary α-β coordinate system α 、u β ; The Park inverse transform formula is:

[0058]

[0059] will u α 、u β The SVPWM module is input to the space vector modulation module, and after the calculation is completed, 6 PWM signals are output; the 6 PWM signals are input to the three-phase inverter, and the inverter converts the DC voltage into a three-phase voltage u a 、u b 、u c The rotation of the permanent magnet synchronous motor is controlled, thereby completing the vector control of the permanent magnet synchronous motor.

[0060] The principle of the integral compensation formula in the present invention is derived as follows:

[0061] The controller model based on deadbeat current prediction is:

[0062] u(k)=G -1 [x * (k+1)-F(k)x(k)-H(k)]

[0063] where x(k) = [i d (k) i q (k)] T ; u(k)=[u d (k) u q (k)] T

[0064]

[0065] where u d (k),u q (k) are the d-axis and q-axis control voltages of the motor in the kth control cycle; Respectively represent the d-axis and q-axis command currents of the motor at the kth moment; i d (k), i q (k) represents the actual current of the motor d-axis and q-axis at the kth moment; R is the stator equivalent resistance, L is the stator equivalent inductance, and T is the motor torque; ψ f is the permanent magnet flux of the motor.

[0066] Since the permanent magnet synchronous motor current loop control period Ts is usually around 100us, and the motor stator resistance R is generally several hundred milliohms, and the inductance L is generally less than 20mH, the mismatch of the resistance R has little effect on the overall control performance of the current loop for the (1-TR / L) term in the formula. Here, (1-TR / L) can be approximately equal to 1, and the above formula can be simplified to:

[0067] u(k)=G -1 [x * (k+1)-I(k)x(k)-H(k)]

[0068] in

[0069] Assume that the actual values ​​of the motor inductance and permanent magnet flux in the actual system are L0 and ψ f0 , then the actual current prediction result is as follows:

[0070] x(k+1)=I(k)x(k)+G0u(k)+H0(k)

[0071] in

[0072] Combining the two equations, we get the relationship between the command current and the response current when there are errors in the motor inductance and permanent magnet flux parameters:

[0073]

[0074] where ∆ψ f =ψ f -ψ f0 is the difference between the flux value in the mathematical model and the motor flux value in the actual system, and △L=L-L0 is the error between the inductance in the mathematical model and the inductance parameter in the actual system.

[0075] When the system is running in steady state, since the time interval between two consecutive control cycles is very small and the current change is also very small, it can be considered that id (k+1)=i d (k), i q (k+1)=i q (k), the above formula is simplified to:

[0076]

[0077] Combining the above formula, it can be seen that the error between the actual motor parameters and the estimated motor parameters in the deadbeat current prediction control model will cause a static error between the given reference current and the feedback current of the dq axis. When the estimated inductance parameter value in the prediction model is smaller than the actual motor inductance parameter value, △L<0, the actual value of the d-axis current will be greater than the reference value, that is, Similarly, when the estimated inductance parameter value in the prediction model is greater than the actual motor inductance parameter value, △L>0, the actual value of the d-axis circuit will be less than the reference value, that is, Since in the q-axis current prediction equation, the change of motor inductance parameters is (△Li d (k)+△ψ f )Tω e The influence of (k) / L0 on i is very small from the perspective of order of magnitude, so the effect of motor inductance parameter error on i can be ignored. q The influence of current static difference.

[0078] The error of permanent magnet flux parameter in the control model will also lead to the static error problem of dq axis current. When the permanent magnet flux parameter value estimated in the prediction model is less than the actual motor flux parameter value, △ψ f <0, the actual value of the q-axis current will be smaller than the reference value, that is, i q q * Similarly, when the permanent magnet flux parameter value estimated in the prediction model is greater than the actual motor flux parameter value, △ψ f >0, the actual value of the q-axis current will be greater than the reference value, that is, i q >i q * From formula (4-14), it can be seen that the permanent magnet flux error has no direct effect on the d-axis current, but its effect on the q-axis current will indirectly affect the i d , but from the point of view of order of magnitude, the permanent magnet flux error has a great influence on i d The effect of the current static difference can be ignored.

[0079] After the above analysis, the current static error problem caused by the inductance parameter error can be solved by introducing the d-axis current error integral into the d-axis voltage result in view of the inductance parameter error. The integral compensation formula is as follows:

[0080]

[0081] ​Consider the q-axis current static error caused by the permanent magnet flux error separately, and combine it with i d =0, we get:

[0082]

[0083] It can be known that q The current error is proportional to the flux error and can be calculated based on i q The error is used to adjust the value of the flux linkage so that the flux linkage parameters converge to the real value in the motor. The integral compensation formula is as follows:

[0084]

[0085] Combined with the above formulas and derivations, the present invention adjusts the d-axis voltage in the control system in real time by increasing the integral of the error between the actual d-axis current and the command current, thereby eliminating the d-axis current static error problem caused by the inductance parameter error; by increasing the integral of the error between the actual q-axis current and the command current, the flux estimation parameter in the control system is corrected to make it close to the true value of the motor flux, thereby fundamentally solving the current static error problem caused by the flux parameter error. The core idea of ​​the integral compensation method is to utilize the zero-difference characteristic of the integral link, introduce an integral link to integrate the current static error, and use the output of the integral link as a feedforward compensation to compensate the d-axis voltage and flux parameters, thereby eliminating the current static error caused by the inductance and flux parameter errors in the system.

[0086] During the actual operation of the motor, the d-axis current command value is generally kept at 0, and the q-axis command current often changes suddenly. Therefore, the method of compensating the flux parameters by integrating the static difference of the q-axis current avoids the problem of q-axis voltage mutation and even system instability caused by directly performing integral compensation on the q-axis voltage. Converging the flux parameter compensation to the true value of the motor flux can fundamentally solve the problems caused by the flux parameter error in the zero-beat current prediction model.

[0087] In order to verify the reliability of the present invention, relevant simulation experiments were carried out, and the test motor was controlled by the current loop. Figure 2 (a) and (b) are the traditional PI current loop and the motor parameter accurate deadbeat current prediction control current loop under a given command current i q * From the simulation results under the conditions of increasing from 1A to 2A at 0.1s and then decreasing from 2A to 0.5A at 0.3s, it can be seen that under ideal conditions, the dynamic and static performance of the deadbeat current prediction controller are better than that of the PI current control. Figure 3(a) and (b) are the simulation results before and after integral compensation after the inductor parameter error is introduced into the deadbeat current prediction control algorithm. It can be seen that the inductor parameter error will cause the d-axis current static error problem, and the integral compensation can eliminate the d-axis current static error, which is consistent with the above analysis results. Figure 4 (a) and (b) are the simulation results before and after the integral compensation is added after the deadbeat current predictive control algorithm introduces the permanent magnet flux parameter error. It can be seen that when there is an error in the flux, it will bring obvious current static error to the q-axis current. After introducing integral compensation, the q-axis current static error can be eliminated, which is also consistent with the analysis results.

[0088] The method proposed in the present invention can retain the advantages of high bandwidth and simple calculation of the deadbeat current predictive control algorithm, while solving the problem of dq-axis current static error caused by motor parameter errors. Therefore, compared with the traditional PI-based current loop system, the present invention has better dynamic response performance, simple calculation, and is suitable for engineering application.

[0089] The present invention also provides a deadbeat current prediction control system based on integral compensation. The functions of each module in the system correspond to the steps of the method, and the present invention will not be repeated here.

[0090] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A deadbeat current prediction control method based on integral compensation, characterized in that: include: S1. According to the actual value of the motor q-axis current i q With the instruction value i q * The error between the two is used to perform integral compensation on the permanent magnet flux parameters of the motor. The following formula is used to perform integral compensation on the permanent magnet flux parameters of the motor: ψ f com is the permanent magnet flux of the motor after integral compensation, ψ f is the permanent magnet flux of the motor, The q-axis current command value of the motor at the kth moment, i q (k) is the actual value of the motor's q-axis current at the kth moment, K iψ is the flux linkage integral coefficient; S2. Take the motor d-axis current command value, q-axis current command value and the motor permanent magnet flux parameters after integral compensation as input, perform deadbeat current predictive control on the motor, and obtain the motor d and q axis control voltage u d 、u q ; The controller model based on deadbeat current prediction is: u(k)=G -1 [x * (k+1)-F(k)x(k)-H(k)] where x(k) = [i d (k) i q (k)] T ; u(k) = [u d (k) u q (k)] T where u d (k),u q (k) are the d-axis and q-axis control voltages of the motor in the kth control cycle; Respectively represent the d-axis and q-axis command currents of the motor at the kth moment; i d (k), i q (k) represents the actual current of the motor d-axis and q-axis at the kth moment; R is the stator equivalent resistance, L is the stator equivalent inductance, and T is the motor torque; ψ f is the permanent magnet flux of the motor; S3. According to the actual value of motor d-axis current i d With the instruction value i d * The error between them controls the motor d-axis voltage u d Perform integral compensation; use the following formula to control the motor d-axis voltage u d To make points compensation: u d (k) com is the d-axis control voltage of the motor in the kth control cycle after integral compensation, u d (k) is the d-axis control voltage of the motor in the kth control cycle, The d-axis current command value of the motor at the kth moment, i d (k) is the actual value of the motor's d-axis current at the kth moment, K id is the d-axis voltage integral coefficient; S4. Use the motor q axis to control the voltage u q And the motor d-axis control voltage u after integral compensation d The motor is pulse-width modulated using space vector technology.

2. The deadbeat current prediction control method based on integral compensation according to claim 1, characterized in that: Actual value of motor q-axis current i q With the instruction value i q * The error between and the permanent magnet flux parameter error satisfies the following relationship: L0 is the actual value of the inductance parameter, T is the motor torque, ω e (k) is the Δψ of the motor at the kth moment f is the permanent magnet flux parameter error.

3. A deadbeat current prediction control system based on integral compensation, characterized in that: include: Flux compensation module, deadbeat current prediction controller, d-axis control voltage compensation module and SVPWM modulation module; The flux compensation module is used to compensate the actual value of the motor q-axis current i q With the instruction value i q * The error between the two is used to perform integral compensation on the permanent magnet flux parameters of the motor. The following formula is used to perform integral compensation on the permanent magnet flux parameters of the motor: ψ f com is the permanent magnet flux of the motor after integral compensation, ψ f is the permanent magnet flux of the motor, The q-axis current command value of the motor at the kth moment, i q (k) is the actual value of the motor's q-axis current at the kth moment, K iψ is the flux linkage integral coefficient; The deadbeat current prediction controller is used to perform deadbeat current prediction control on the motor with the motor d-axis current command value, q-axis current command value and the motor permanent magnet flux linkage parameter after integral compensation as input to obtain the motor d-axis and q-axis control voltage u d 、u q ; The controller model based on deadbeat current prediction is: u(k)=G -1 [x * (k+1)-F(k)x(k)-H(k)] where \(x(k)=[i d (k)i q (k)] T ; \(u(k)=[u d (k)u q (k)] T where u d (k),u q (k) are the d-axis and q-axis control voltages of the motor in the kth control cycle; Respectively represent the d-axis and q-axis command currents of the motor at the kth moment; i d (k), i q (k) represents the actual current of the motor d-axis and q-axis at the kth moment; R is the stator equivalent resistance, L is the stator equivalent inductance, and T is the motor torque; ψ f is the permanent magnet flux of the motor; The d-axis control voltage compensation module is used to adjust the actual value of the motor d-axis current i d With the instruction value i d * The error between them controls the motor d-axis voltage u d Perform integral compensation; use the following formula to control the motor d-axis voltage u d To make points compensation: u d (k) com is the d-axis control voltage of the motor in the kth control cycle after integral compensation, u d (k) is the d-axis control voltage of the motor in the kth control cycle, The d-axis current command value of the motor at the kth moment, i d (k) is the actual value of the motor's d-axis current at the kth moment, K id is the d-axis voltage integral coefficient; SVPWM modulation module, used to control the voltage u using the motor q axis q And the motor d-axis control voltage u after integral compensation d The motor is pulse-width modulated using space vector technology.

Images