Low speed synchronous machine sensorless control method with rotating square wave voltage injection

Abstract

The application relates to a low-speed synchronous motor position sensorless control method of rotating square wave voltage injection, and relates to the technical field of motor control. The application is aimed at solving the problems of the existing position sensorless control method, i.e. the need of motor parameters for parameter setting, the introduction of a filter limiting the system bandwidth and leading to the decline of the dynamic performance of the system, and the reduction of the dynamic performance by the ellipse fitting technology. The application injects rotating square wave voltage into the gamma-delta axis of a low-speed synchronous reluctance motor, obtains high-frequency induced current of the alpha-beta axis, and constructs an ellipse trajectory equation based on the high-frequency induced current; the high-frequency induced current is reconstructed, the current ellipse trajectory distortion is compensated by using the reconstructed current; the least square method is applied to estimate the coefficients of the ellipse trajectory equation; the sine-cosine angle processing is performed on the angle between the d e q e And the alpha-beta coordinate system; the Q-PLL is adopted to estimate the rotor position and speed, and the estimated value of the rotor position and speed is used to realize the position sensorless control of the low-speed synchronous reluctance motor.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of motor control. BACKGROUND

[0002] Synchronous reluctance motor (SynRM) has the characteristics of simple structure, low cost and strong overload capacity, and has been widely used in industrial drive and electric vehicle application fields in recent years. Accurate rotor position is the key to achieving high-performance synchronous motor control. However, the application of position sensors increases the cost, so it is of practical significance to study the position sensorless control method.

[0003] Current position sensorless control methods mainly include observer-based methods and high-frequency injection-based methods. Observer-based methods obtain position information by estimating stator flux or extended EEMF. However, stator flux estimation is affected by insufficient signal-to-noise ratio at low motor speed, and the amplitude of EEMF is proportional to rotor speed, so the observer-based method is suitable for medium and high speed regions.

[0004] In order to expand the speed range of position sensorless control, high-frequency injection-based methods are proposed. High-frequency injection-based methods extract rotor position by observing high-frequency induced current caused by high-frequency injection signal through an observer. However, the demodulation of high-frequency current inevitably introduces a specific filter, which complicates the system and reduces the dynamic performance. In addition, in order to obtain excellent control performance, the observer parameters need to be accurately designed according to the motor parameters. However, when the motor parameters change, especially when the inductance parameters change due to magnetic saturation, the observer parameters are mismatched, causing the system bandwidth to change and reducing the dynamic performance of the controller.

[0005] In recent years, ellipse fitting technology has been applied to position sensorless control due to its easy parameter tuning and no need for motor parameters. This method injects a rotating voltage vector into the stator winding and estimates the rotor position according to the spatial relationship between the long semi-axis of the elliptical high-frequency induced current trajectory and the direct axis of the rotor. In order to extract the high-frequency elliptical current trajectory, the current method needs to introduce a high-pass filter in the demodulation of the signal, which reduces the dynamic performance of the control system.

[0006] Based on the introduction of the above background technology, the current position sensorless control method needs motor parameters for parameter tuning, and the introduction of the filter limits the system bandwidth, resulting in a decrease in the dynamic performance of the system. Although the existing ellipse fitting technology avoids the need for motor parameters for tuning the controller, it does not introduce a filter, which requires the use of an ellipse equation with a center not at the origin, the equation is relatively complex, and the parameter estimation calculation is large. The introduction of the filter simplifies the equation, but reduces the dynamic performance. SUMMARY

[0007] This invention addresses the problems of existing sensorless control methods requiring motor parameters for tuning, the introduction of filters limiting system bandwidth and leading to a decrease in dynamic performance, and ellipse fitting techniques further reducing dynamic performance. The invention provides a sensorless control method for low-speed synchronous reluctance motors based on rotating square wave voltage injection.

[0008] A sensorless control method for low-speed synchronous reluctance motors based on rotating square wave voltage injection includes:

[0009] A rotating square wave voltage is injected into the γδ axis of a low-speed synchronous reluctance motor to obtain a high-frequency induced current on the αβ axis, and an elliptical trajectory equation is constructed based on this high-frequency induced current.

[0010] The high-frequency induced current of the αβ axis is reconstructed, and the reconstructed current signal is used to compensate for the distortion of the current elliptical trajectory caused by the non-zero velocity in the historical period and the current period.

[0011] Based on the compensated current, the coefficients of the elliptical trajectory equation are estimated using the least squares method;

[0012] Using the coefficient estimates of the elliptic trajectory equation for d e q e Angle with the αβ coordinate system Perform sine and cosine angle processing, the d e q e The coordinate system is a magnetically convex polar coordinate system;

[0013] Based on the included angle and d-axis and d e The angle θ between the axes m Q-PLL is used to estimate the rotor position and speed, and the estimated rotor position and speed values ​​are used to achieve sensorless control of low-speed synchronous reluctance motor.

[0014] Furthermore, the above-mentioned injection of a rotating square wave voltage into the γδ axis of the low-speed synchronous reluctance motor to obtain a high-frequency induced current in the αβ axis includes:

[0015] The expression for the rotating square wave voltage is:

[0016]

[0017] Among them, u αh and u βh The high-frequency injection voltages for the α and β axes are respectively, where n is the voltage injection period and ω is the high-frequency injection voltage for the α and β axes. inj T is the angular velocity of rotation along the γδ axis. s Let T(·) be the PWM signal period, and T(·) be the coordinate transformation matrix.γδh a rotating square-wave voltage vector is injected to the γδ axis, U h is the amplitude of the rotating square-wave voltage injected to the γδ axis, and k is a control period;

[0018] The expression of the high-frequency induced current of the αβ axis is:

[0019]

[0020] where p is a differential operator, i αh and i βh are the high-frequency induced currents of the α and β axes, respectively, and are the differential inductances of the d e and q e axes, respectively, θ inj is the included angle between the α axis and the γ axis, θ e is the included angle between the α axis and the d e axis, l Σ is the average value of the dq-axis inductance, and l Δ is the half-difference value of the dq-axis inductance.

[0021] Further, the above constructing the elliptical trajectory equation comprises:

[0022] The expression of the elliptical trajectory equation is:

[0023] A(pi αh ) 2 +Bpi αh pi βh +C(pi βh ) 2 = 1,

[0024] where A, B, and C are all coefficients of the elliptical trajectory equation.

[0025] Further, the above reconstructing the high-frequency induced current of the αβ axis comprises:

[0026] The reconstruction equation of the high-frequency induced current signal is:

[0027]

[0028] where I αβh is the reconstructed high-frequency current vector of the αβ axis, i αβ is the sampled current vector of the αβ axis, i αβh is the high-frequency induced current vector of the αβ axis before reconstruction, n is a voltage injection period, and p is a differential operator.

[0029] The reconstruction equation of the fundamental frequency current signal is:

[0030]

[0031] wherein I αβb is the reconstructed fundamental current vector of αβ axis, k is the control period.

[0032] Further, the above-mentioned compensation for the current ellipse trajectory distortion caused by non-zero speed in the historical period and the current period by using the reconstructed current signal includes:

[0033] The non-zero speed distortion compensation method for the current data in the historical voltage injection period is:

[0034]

[0035] wherein, is the compensated current vector of αβ axis, is the estimated speed of the nth voltage injection period, T(·) is the coordinate transformation matrix, T s is the PWM signal period;

[0036] The non-zero speed distortion compensation method for the current data in the current voltage injection period is:

[0037]

[0038] Further, the above-mentioned estimation of the coefficients of the ellipse trajectory equation by using the compensated current by using the least square method includes:

[0039] The expression of the least square method for estimating the coefficients of the ellipse trajectory equation is:

[0040] y = ΦΘ,

[0041] wherein,

[0042] and are the compensated currents of α and β axes respectively, A, B, C are all the coefficients of the ellipse trajectory equation, and N is the total number of voltage injection periods; the estimated value of the coefficients of the ellipse trajectory equation is:

[0043]

[0044] are the estimated values of A, B, and C respectively.

[0045] Further, the above-mentioned d e q e is the angle between the αβ coordinate system and the d are processed by using the sine and cosine angle, including:

[0046] Using the coefficient estimates of the elliptic trajectory equation for d e q e Angle with the αβ coordinate system The expression for sine and cosine angle processing is:

[0047]

[0048] These are the estimated values ​​of the coefficients A, B, and C of the elliptical trajectory equation, respectively.

[0049] Furthermore, the expressions for the coefficients A, B, and C of the above trajectory equation are as follows:

[0050]

[0051] in,

[0052] Furthermore, the above is based on the included angle and d-axis and d e The angle θ between the axes m The rotor position and speed are estimated using Q-PLL, including:

[0053] Rotor position estimation The expression is:

[0054]

[0055] Where k is the control period, i d and i q The d-axis and q-axis currents are pre-calculated offline and stored as lookup table values;

[0056] Rotor speed estimate The expression is:

[0057]

[0058] Among them, T s The PWM signal period.

[0059] Furthermore, the function expression for the above Q-PLL is:

[0060]

[0061] Where, k p =2ξω n , ω is the damping coefficient. n s is the bandwidth of the Q-PLL, and s is the Laplace operator.

[0062] The low-speed synchronous reluctance motor position sensorless control method based on the rotating square wave voltage injection has the following beneficial effects:

[0063] 1. The high-frequency rotating voltage injection adopts a square wave voltage form, and the high-frequency induced current signal is demodulated by using a current vector reconstruction mode, compared with a traditional signal processing mode using a filter, the control system complexity is not increased, and the dynamic performance of the system is improved.

[0064] 2. The method has the advantages of easy parameter setting and not requiring motor parameters. The method compensates for the distortion of the elliptical trajectory caused by the non-zero speed, and accurate motor angle position estimation can be obtained at different speeds.

[0065] 3. The position error caused by magnetic saturation is compensated by using a pre-calculated lookup table, a Q-PLL is used to obtain smooth rotor position estimation and speed estimation, the position accuracy is obviously improved, and the angle position estimation error does not increase with the increase of the load. BRIEF DESCRIPTION OF DRAWINGS

[0066] Figure 1 is the high-frequency induced current in the alpha-beta coordinate system when the motor is at rest;

[0067] Figure 2 is a timing diagram of the high-frequency rotating square wave voltage injection and sampling;

[0068] Figure 3 is a signal demodulation block diagram, (a) represents high-frequency induced current extraction, and (b) represents fundamental current extraction;

[0069] Figure 4 is an elliptical fitting schematic diagram of the high-frequency induced current with compensation, (a) represents zero speed, and (b) represents non-zero speed;

[0070] Figure 5 is an elliptical fitting flowchart with NZSD effect compensation;

[0071] Figure 6 is a Q-PLL principle diagram with magnetic saturation compensation function;

[0072] Figure 7 is a sensorless control block diagram based on high-frequency rotating square wave voltage injection;

[0073] Figure 8 is a result schematic diagram under variable speed conditions, (a) represents no NZSD effect compensation, and (b) represents NZSD effect compensation;

[0074] Figure 9 is a result schematic diagram under load change conditions, (a) represents no magnetic saturation compensation, and (b) represents magnetic saturation compensation. DETAILED DESCRIPTION

[0075] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative effort belong to the scope of protection of the present application. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0076] Reference Figures 1 to 5 To specifically describe the present embodiment, the low-speed synchronous reluctance motor position sensorless control method based on rotating square wave voltage injection in the present embodiment comprises:

[0077] Step 1: injecting rotating square wave voltage into γδ axis rotating at a specific angular frequency to obtain high-frequency induced current data of αβ axis; introducing magnetic salient pole coordinate system d e q e , and deriving current ellipse trajectory equation. Specifically as follows:

[0078] Combining Figure 2 It can be known that the specific expression of the injected rotating square wave voltage equation is:

[0079]

[0080] Among them, u αh and u βh are high-frequency injected voltages of α and β axes respectively, ω inj is the angular velocity of γδ axis rotation, U h is the amplitude of the injected rotating square wave voltage of γδ axis, k is the control period, n is the number of voltage injection periods, represents the coordinate transformation matrix, u γδh represents the rotating square wave voltage vector injected into the γδ axis, T s is the PWM signal period.

[0081] Introducing magnetic salient pole coordinate system d e q e , the expression of high-frequency induced current data of αβ axis is:

[0082]

[0083] Among them, p is a differential operator, i αh and i βh are high-frequency induced currents of α and β axes respectively, and are d e , qe The differential inductance of the shaft, θ inj θ is the angle between the α-axis and the γ-axis. e Let α be the axis and d e The angle between axes. This represents the average inductance along the dq axis. This represents the half-dq axis inductance difference.

[0084] d e q e The relationship between the differential inductance of the d-axis and the differential inductance of the q-axis is as follows:

[0085]

[0086] Among them, l d and l q Differential inductances along the d and q axes, respectively, l dq Let θ be the cross-saturation inductance along the dq axis. m d-axis and d e The angle between axes (according to i) d i q These are the pre-calculated and stored lookup table values ​​for the d-axis and q-axis currents, respectively.

[0087] Combination Figure 1 It can be seen that the trajectory equation of a high-frequency current is an ellipse, and the expression for this elliptical trajectory equation is:

[0088] A(pi αh ) 2 +Bpi αh pi βh +C(pi βh ) 2 =1,

[0089] The elliptic coefficients A, B, and C are represented as follows:

[0090]

[0091]

[0092] Step 2 involves demodulating the high-frequency induced current data obtained in Step 1, mainly by reconstructing the high-frequency current signal and the fundamental frequency current signal; and compensating for the distortion of the current elliptical trajectory caused by the non-zero velocity in the historical period and the current period, to obtain the compensated current sample.

[0093] Combination Figure 3 It can be seen that the reconstruction equation for the high-frequency induced current signal is:

[0094]

[0095] Iαβh is the high frequency current vector reconstructed for αβ axis, i αβ is the sampled current vector for αβ axis, i αβh is the high frequency induced current for αβ axis.

[0096] The reconstruction equation for fundamental current signal is:

[0097]

[0098] I αβb is the fundamental current for αβ axis.

[0099] Combining Figure 4 It can be seen that the non-zero speed distortion will affect the accuracy of position estimation, so the current compensation method for NZSD (non-zero speed distortion) is proposed, as shown in the flowchart of Figure 5

[0100] The non-zero speed distortion compensation method for current data in the historical voltage injection period is:

[0101]

[0102] wherein, is the compensated αβ axis current vector, represents the estimated speed of the nth voltage injection period, n = 1, 2, …, N, and N is the total number of voltage injection periods.

[0103] The non-zero speed distortion compensation method for current data in the current voltage injection period is:

[0104]

[0105] Step 3, according to the compensated current sample obtained in step 2, apply least square method to the elliptical trajectory equation coefficient in step 1.

[0106] The least square method estimates the elliptical trajectory equation coefficient expression:

[0107] y = ΦΘ,

[0108] wherein,

[0109] The estimated value of the elliptical trajectory equation coefficient

[0110] Step 4, the sine and cosine angle processing is performed on , and then the θ m The rotor position and speed are estimated, and the sine and cosine angle processing in this step can obtain smooth rotor position and rotor speed estimation value.

[0111] ​The d e q e The angle between the αβ coordinate system and the d The expression for performing the sine and cosine angle processing is:

[0112]

[0113] The rotor position and speed are estimated by using a Q-PLL (phase-locked loop):

[0114] Rotor position estimate

[0115]

[0116] Rotor speed estimate

[0117]

[0118] The expression of the Q-PLL is:

[0119]

[0120] wherein k p = 2ξω n , is a damping coefficient, ω n is the bandwidth of the Q-PLL, and s is a Laplace operator. Embodiment

[0122] To further illustrate the effectiveness of the method and the specific implementation process thereof, a specific implementation case of the present application is given. A 3-kw synchronous reluctance motor is used as the experimental object, the control frequency is 10 kHz, the injected voltage amplitude is 30 V, and the angular frequency of the γδ coordinate system is 1700π.

[0123] Figure 6 The Q-PLL principle diagram with a magnetic saturation compensation function is shown. Smooth position and speed estimates can be obtained by the Q-PLL. Figure 7 The block diagram of the proposed sensorless control based on the rotating square-wave voltage injection is shown.

[0124] In Figure 8 , the load is 100% rated load, and the speed command is changed in steps of 100 rpm from 0 to 300 rpm and then back to 0 rpm. Due to the NZSD effect, the position error increases with the increase of the speed, as shown in Figure 8 (a). At a speed of 300 rpm, the position error can reach 5.51°. After compensating the NZSD effect, the position error is less than 1.21°.

[0125] InFigure 9 In this case, the speed was 100 rpm, and the load was increased stepwise from 25% to 100% and then back to 25% rated load. Figure 9 (a) shows that magnetic saturation causes the position error to increase with increasing load and causes the motor to lose control at 100% load. After compensation, as shown in Figure 9 (b), the position accuracy is significantly improved, and the motor operates stably at rated load.

[0126] While the application has been described with reference to particular embodiments, it will be understood that the examples are merely illustrative of the principles and applications of the application. It will be understood that numerous modifications can be made to the illustrative examples and that other arrangements can be devised without departing from the spirit and scope of the present application as defined by the appended claims. It will be understood that the features of the various embodiments can be combined with each other, in different ways than as described herein. It will be understood that features described with reference to individual embodiments can be used in other described embodiments.

Claims

1. A low speed synchronous machine sensorless control method of rotating square wave voltage injection, characterized in that, The application relates to a method for estimating the parameters of an ellipse trajectory equation of a low-speed synchronous reluctance motor. A rotating square wave voltage is injected into the gamma-delta axis of the low-speed synchronous reluctance motor to obtain high-frequency induced current of the alpha-beta axis, and an ellipse trajectory equation is constructed based on the high-frequency induced current; The high-frequency induced current of the alpha-beta axis is reconstructed, and the reconstructed current signal is used to compensate for the distortion of the current ellipse trajectory caused by non-zero speed in the historical period and the current period; The least square method is applied to estimate the coefficients of the ellipse trajectory equation according to the compensated current; Using the elliptical trajectory equation coefficient estimate value for d e q e The angle between the coordinate system and the αβ coordinate system Performing sine and cosine angle processing, the d e q e The coordinate system is a magnetic salient pole coordinate system Based on the included angle And the included angle θ between the d-axis and the d-axis e m The rotor position and speed are estimated by using Q-PLL, and the estimated value of the rotor position and speed is used to realize the sensorless control of the low-speed synchronous reluctance motor.​ The reconstruction of the high-frequency induced current of the alpha-beta axis comprises the following steps: The reconstruction equation of the high-frequency induced current signal is as follows: where I αβh is the high-frequency current vector reconstructed for the αβ axis, i αβ is the sampled current vector for the αβ axis, i αβh is the high-frequency induced current vector before reconstruction for the αβ axis, n is the voltage injection period, and p is the differentiation operator; The reconstruction equation of the fundamental frequency current signal is as follows: where I αβb is the fundamental current vector reconstructed for the αβ axes, and k is the control period. The compensation of the current ellipse trajectory distortion caused by non-zero speed in the historical period and the current period by using the reconstructed current signal comprises the following steps: The non-zero speed distortion compensation method for the current data in the historical voltage injection period is as follows: wherein, is the compensated αβ-axis current vector, is the estimated velocity for the nth voltage injection period, T(·) is the coordinate transformation matrix, T s is the PWM signal period; The non-zero speed distortion compensation method for the current data in the current voltage injection period is as follows:

2. The low speed synchronous machine position sensorless control method of field-oriented control with rotary square-wave voltage injection according to claim 1, characterized in that, The injection of the rotating square wave voltage into the gamma-delta axis of the low-speed synchronous reluctance motor to obtain the high-frequency induced current of the alpha-beta axis comprises the following steps: The expression of the rotating square wave voltage is as follows: wherein u αh and u βh are high-frequency injection voltages of the α and β axes, respectively, n is a voltage injection period, ω inj is a γδ-axis rotational angular velocity, T s is a PWM signal period, T(·) is a coordinate transformation matrix, u γδh is a rotational square-wave voltage vector injected into the γδ axes, U h is an amplitude of the rotational square-wave voltage injected into the γδ axes, and k is a control period; The expression of the high-frequency induced current of the alpha-beta axis is as follows: where p is a differential operator, i αh and i βh are high-frequency induction currents of the α and β axes, respectively, and are differential inductances of the d e and q e axes, respectively, θ inj is an angle between the α axis and the γ axis, θ e is an angle between the α axis and the d e axis, l Σ is an average value of the dq-axis inductance, l V is a half-difference value of the dq-axis inductance, and 3. The low speed synchronous machine position sensorless control method of field-oriented control with rotary square-wave voltage injection according to claim 2, characterized in that, The construction of the ellipse trajectory equation comprises the following steps: The expression of the ellipse trajectory equation is as follows: A(pi αh ) 2 +Bpi αh pi βh +C(pi βh ) 2 =1, Wherein, A, B and C are all the coefficients of the ellipse trajectory equation.

4. The low speed synchronous machine position sensorless control method of field-oriented control with rotary square-wave voltage injection of claim 1, wherein, The estimation of the coefficients of the ellipse trajectory equation by applying the least square method according to the compensated current comprises the following steps: The expression of the least square method for estimating the coefficients of the ellipse trajectory equation is as follows: y=Phi Theta, wherein, and are compensated α and β axis currents, respectively, A, B, C are all elliptical trajectory equation coefficients, and N is the total number of voltage injection periods; the elliptical trajectory equation coefficient estimation value is: Estimates of A, B, C, respectively.

5. The low speed synchronous machine position sensorless control method of field-oriented control with rotary square-wave voltage injection as claimed in claim 1, characterized in that, The d e q e The angle between the coordinate system and the αβ coordinate system The sine and cosine angle processing includes: Using the coefficient estimates of the elliptical trajectory equation for d e q e The angle between the coordinate system and the αβ coordinate system The expression for performing the sine-cosine angle processing is: respectively the estimated values of the coefficients A, B, C of the elliptic trajectory equation.

6. The low speed synchronous machine position sensorless control method of field oriented control with rotary square wave voltage injection of claim 3, wherein, The expression of the trajectory equation coefficients A, B and C is as follows: wherein 7. The rotary low speed synchronous machine position sensorless control method of square-wave voltage injection of claim 1, wherein, The angle-based and the angle θ between the d-axis and the d e axis m Q-PLL is used to estimate the rotor position and speed, comprising: Rotor position estimate The expression is: where k is a control period, i d and i q are d and q axis current pre-off-line calculated stored look-up values, respectively; Rotor speed estimate The expression is: where T s is the PWM signal period.

8. The low speed synchronous machine position sensorless control method of field-oriented control with rotary square-wave voltage injection according to claim 7, characterized in that, The function expression of the Q-PLL is as follows: where k p = 2ξω n , ξ is a damping coefficient, ω n is the bandwidth of the Q-PLL, and s is the Laplace operator.

Images