Model reference adaptive permanent magnet synchronous motor speed sensorless control method

Abstract

The application discloses a model reference adaptive permanent magnet synchronous motor speed sensorless control method, aiming at the problem that a flux linkage observer of a second-order generalized integrator (SOGI) cannot accurately observe the motor speed, a model reference adaptive (MRAS) is introduced to replace a frequency lock loop (FLL) module in the flux linkage observer, a nonlinear error feedback system is constructed by a reference model and an adjustable model to provide a stable center frequency for the SOGI, and an MRAS adaptive law is designed based on a super-spiral sliding mode control method, a sigmoid (s) function with smooth continuous characteristics is adopted to reduce inherent chattering of a sliding mode and improve observation accuracy of a rotor position and a rotor speed of the permanent magnet synchronous motor.Combined with a vector control technology, the model reference adaptive control flux linkage observer (MSOGI) can effectively improve dynamic performance and steady precision of a speed sensorless control system of the permanent magnet synchronous motor and widen a speed regulating range of the motor.

Description

Technical Field

[0001] This invention relates to the field of motor control technology, specifically to a sensorless control method for a model reference adaptive permanent magnet synchronous motor. Background Technology

[0002] Permanent magnet synchronous motors (PMSMs) are widely used in industry due to their advantages such as small size, high efficiency, low energy consumption, and reliable operation. High-performance control of PMSMs relies on accurate rotor position information. However, mechanical sensors increase system cost and device size, and their accuracy is affected by external environmental factors. Therefore, sensorless control technology has attracted widespread attention. Currently, sensorless control technologies are mainly divided into two categories: one is the high-frequency signal injection method, which is only applicable to zero speed and low speed; the other is the observer method, mainly including sliding mode observers, extended Kalman filters, and flux linkage observers. Among them, the flux linkage observer obtains the rotor flux position and angle by integrating the back electromotive force (EMF) of the motor, and has gained widespread application. However, due to factors such as unknown initial integral values, parameter mismatch, measurement errors, noise, and inverter dead zones in actual systems, the EMF introduces DC bias and high-order harmonic components, leading to a decrease in the observation accuracy of the flux linkage observer. Second-order generalized integrators not only possess excellent integration characteristics but also exhibit good filtering properties against DC bias and higher harmonic components, thus they are frequently used in sensorless control of PMSMs. The implementation of a second-order generalized integrator requires the center frequency of the input signal, typically obtained through a frequency-locked loop (LLL) module. However, the center frequency of the input signal is also affected by higher harmonics and DC bias, especially at low motor speeds, where the LLL module may experience center frequency fluctuations or even loss of lock-in.

[0003] To address the problems of second-order generalized flux observers, the published paper "An Improved Third-Order Generalized Integral Flux Observer for Sensorless Drive of PMSMs" constructs a third-order generalized flux integrator by adding an integrator to the control loop. The performance of this method mainly depends on the parameter design of the observer. A narrower bandwidth provides better filtering capability and better suppression of harmonic components and DC bias, but the additional integrator slows down the system's dynamic response and makes the system parameter design more complex.

[0004] The published paper "Modified Second-Order Generalized Integrators With Modified Frequency Locked Loop for Fast Harmonics Estimation of Distorted Single-Phase Signals" proposes an improved frequency-locked loop (MFLL). This method, based on the traditional FLL, introduces gain normalization, a generalized adaptive frequency law, a sign correction anti-integral saturation strategy, and a frequency change rate limit, ensuring that the estimated rotor angular frequency remains positive and limited within a set range, thus avoiding observer instability. The improved MFLL offers faster frequency estimation and stronger resistance to harmonic interference, and can be used in conjunction with a parallel modified second-order generalized integrator (SOGI) to achieve rapid estimation of multiple harmonics, improving tracking speed and smoothness during sudden changes in motor speed. However, the system structure and parameter design are more complex than the traditional FLL, requiring careful setting of adaptive gain, frequency upper and lower limits, and rate-of-change limit parameters. Inappropriate parameter selection may still affect dynamic response speed and estimation accuracy.

[0005] Therefore, effectively suppressing the influence of DC bias and high-order harmonic components on the flux linkage observer, improving the observation accuracy and speed of the flux linkage observer on the PMSM rotor flux linkage, and enhancing the control performance of the sensorless control system when the motor is running at low speed are of great significance. Summary of the Invention

[0006] The purpose of this invention is to provide a sensorless control method for a model reference adaptive permanent magnet synchronous motor (PMSM) that enables precise observation of the rotor angle and flux linkage, thereby overcoming the shortcomings of existing technologies.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] The sensorless control method for model-reference adaptive permanent magnet synchronous motors includes the following steps:

[0009] S1: Based on the current equations in the synchronous rotating coordinate system, establish the reference model and the adjustable model to obtain the stator current. , and its estimated value , ;

[0010] S2: According to , , , A nonlinear time-varying error feedback system is constructed, and the rotor flux electric angular velocity is designed using Popov's superstability theorem and the super-helical sliding mode control method. The adaptive law uses a sigmoid(s) function with smooth and continuous characteristics instead of the switching function in traditional sliding mode control. Reduce inherent chattering in sliding mode;

[0011] S3: Calculation - Motor rotor electromotive force in coordinate system The input is then processed by SOGI, and the rotor flux electric angular velocity provided by S2 is used for integration. Using the center frequency of SOGI, it allows for the acquisition of accurate rotor flux estimation values. Calculate the real-time position and speed of the rotor of the permanent magnet synchronous motor;

[0012] S4: Based on vector control technology, the estimated rotor flux calculated in S3. By performing flux orientation and using coordinate transformation, the stator current of the permanent magnet synchronous motor is decoupled and closed-loop controlled in a rotating coordinate system, thereby achieving sensorless control of the permanent magnet synchronous motor.

[0013] Furthermore, the specific method in S1 is to establish a reference model (1) with the actual electric angular velocity as a parameter and an adjustable model (2) with the estimated electric angular velocity as a parameter based on the stator current equation of the permanent magnet synchronous motor in the synchronous rotating coordinate system, as follows:

[0014] (1)

[0015] (2)

[0016] In the formula, , These represent the stator currents along the d-axis and q-axis, respectively. , These represent the stator voltages along the d-axis and q-axis, respectively. Indicates stator inductance; Indicates stator resistance; Indicates the magnetic flux linkage of the rotor permanent magnet; This represents the electric angular velocity of the rotor flux linkage; , These represent the estimated stator current values ​​for the d-axis and q-axis, respectively. This represents an estimated value of the rotor flux linkage electric angular velocity; It represents the differential variable.

[0017] Furthermore, the method for constructing a nonlinear time-varying error feedback system in S2 is as follows: Subtract the adjustable model from the reference model to obtain equation (3):

[0018] (3)

[0019] In the formula, For the stator current estimation error, A= W is the nonlinear feedback channel matrix, and C is the forward channel gain matrix. The output signal of the forward channel is obtained by weighting and synthesizing the stator current estimation error vector e with the forward channel gain matrix C.

[0020] Furthermore, in S2, the sliding surface is selected based on Popov's hyperstability theorem:

[0021] (4)

[0022] Differentiating with respect to the sliding surface, we get:

[0023] (5)

[0024] when At this time, the error feedback system enters the sliding mode and remains on the sliding surface. As defined by the sliding surface, the adjustable model of the motor tends to the reference model, and the electric angular velocity of the motor rotor tends to the actual electric angular velocity, thus achieving accurate observation of the rotor speed and position. Based on the selected sliding surface and the basic form of super-helical sliding mode control, the center frequency is designed, and a control function that continuously switches at zero is adopted. Replace the switching control function that is discontinuous at the zero point. The center frequency of SOGI can be obtained as follows:

[0025] (6)

[0026] In the formula, k1 and k2 are the superspiral sliding mode control gains.

[0027] Furthermore, the permanent magnet synchronous motor in S3 is... - The equation for the rotor electromotive force of the motor in the coordinate system is:

[0028] (7)

[0029] In the formula, , They are permanent magnet synchronous motors axis, Shaft stator electromotive force; , They are respectively axis, Shaft stator current; , They are respectively axis, Shaft stator current.

[0030] Furthermore, the rotor flux estimate The calculation formula is as follows:

[0031] (8)

[0032] In the formula, , The rotor flux obtained by SOGI is respectively in The estimated value on the axis; k is the SOGI gain coefficient, used to adjust the dynamic response and filtering performance; s is the Laplace operator.

[0033] Compared with the prior art, the beneficial effects of the present invention are:

[0034] The sensorless control method for permanent magnet synchronous motors based on model reference adaptive (SMO) of this invention avoids the inherent double frequency fluctuation of the frequency-locked loop compared to the second-order generalized integral flux observer, thereby improving the anti-harmonic interference capability and rotor position estimation accuracy. Secondly, the frequency signal output by the model adaptive method provides a stable frequency reference for the SOGI, ensuring that the SOGI can still efficiently and accurately extract the fundamental component at low speeds, thus improving the control performance of the flux observer at low speeds. Attached Figure Description

[0035] Figure 1 This is a block diagram of the model reference adaptive flux observer structure proposed in this invention;

[0036] Figure 2 This is a block diagram of a sensorless control method for permanent magnet synchronous motors based on a model reference adaptive flux observer proposed in this invention.

[0037] Figure 3 The equivalent nonlinear feedback system structure diagram of the Model Adaptive System (MRAS) module;

[0038] Figure 4 This is a schematic diagram of the MRAS control principle based on the motor stator current. Detailed Implementation

[0039] 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.

[0040] Please see Figure 1The Model Reference Adaptive Flux Observer (MSOGI) in this embodiment of the invention comprises two parts: a model reference adaptive module and a second-order generalized integrator (SOGI). The motor is in - The rotor electromotive force of the motor in the coordinate system is used as the input signal of SOGI, and the rotor flux electric angular velocity estimated by the model reference adaptive module is used as the input signal. The input signal is processed by a second-order generalized integrator to filter out DC bias and higher harmonic components, and the output is the estimated rotor flux linkage value. This is used to calculate the position angle and speed of the electronic rotor. Among them, the permanent magnet synchronous motor... - The equation for the rotor electromotive force of the motor in the coordinate system is:

[0041] (7)

[0042] In the formula, , They are permanent magnet synchronous motors axis, Shaft stator electromotive force; , They are respectively axis, Shaft stator current; , They are respectively axis, Shaft stator current.

[0043] Rotor flux estimation The calculation formula is as follows:

[0044] (8)

[0045] In the formula, , The rotor flux obtained by SOGI is respectively in The estimated value on the axis; k is the SOGI gain coefficient, used to adjust the dynamic response and filtering performance; s is the Laplace operator. Since the center frequency of SOGI is provided by the model reference adaptive system, there is no need to set up a frequency locking loop module, avoiding the inherent second harmonic fluctuation of the frequency locking loop, and improving the flux linkage observation accuracy and anti-loosening capability under low-speed conditions.

[0046] like Figure 2 As shown, the control system in this embodiment of the invention adopts a speed-current loop dual closed-loop control structure. The outer speed loop compares the reference speed input signal with the rotor estimated speed signal, and the PI controller outputs the reference input i of the inner current loop. q * The inner current loop uses id * =0 control strategy, the system detects the three-phase stator current i a i b, i c Then, after Clark coordinate transformation and Park coordinate transformation, it is converted into stator current i in a rotating coordinate system. d i q The inner current loop compares the reference input i. q * and actual stator i q The PI controller outputs the stator voltage signal u in the rotating coordinate system. d , u q Finally, through inverse Park transformation, and then through SVPWM control, the three-phase inverter drives the motor to rotate.

[0047] Depend on Figure 2 It is known that sensorless control systems require a flux linkage observer to provide real-time rotor position and speed. In actual motor control, factors such as inverter nonlinearity, uneven bias in the sampling circuit, dead time of switching devices, measurement errors, and noise can influence the motor's back electromotive force (EMF), introducing DC bias and higher harmonic components, leading to a decrease in the observation accuracy of the flux linkage observer. Second-order generalized integrators not only have good integration characteristics but also good filtering characteristics for DC bias and higher harmonic components, and are therefore often used in sensorless control of PMSMs. The specific implementation of a second-order generalized integrator requires the center frequency of the input signal, usually obtained through a frequency-locked loop module. However, the center frequency of the input signal is also affected by higher harmonics and DC bias, especially at low motor speeds, where the frequency-locked loop module may experience center frequency fluctuations or even loss of lock. This invention utilizes a model reference adaptive module to replace the traditional frequency-locked loop module, providing a model reference adaptive permanent magnet synchronous motor sensorless control method, such as... Figure 3 The diagram shown is an equivalent nonlinear feedback system structure block diagram of the Model Adaptive System (MRAS) module. Essentially, it unifies the reference model, the adjustable model, and the error feedback loop into a closed-loop adaptive structure. It uses current error to drive parameter updates, so that the output of the adjustable model gradually becomes consistent with the output of the reference model, providing a foundation for the subsequent design of the super-helical sliding surface and the construction of the velocity adaptive law.

[0048] like Figure 4As shown, the stator current-based model-reference adaptive system of this invention includes a reference model, an adjustable model, and a super-helical sliding mode control module. The reference model uses the actual electrical angular velocity of the permanent magnet synchronous motor as its model parameter and obtains the stator current based on the stator voltage. The adjustable model uses the estimated electrical angular velocity as its model parameter and obtains an estimated stator current value. The stator current error is obtained by subtracting the stator current estimated value from the stator current estimated value output by the reference model and the stator current estimated value output by the adjustable model. A sliding mode surface is constructed based on the stator current error, and the rotor flux electrical angular velocity estimate is obtained by the super-helical sliding mode control module. The electrical angular velocity estimate is fed back to the adjustable model to make it approach the reference model, and simultaneously output to SOGI as its center frequency. The specific calculation method is as follows:

[0049] Step 1: Establish a reference model and an adjustable model based on the current equation in a synchronously rotating coordinate system to obtain the true current value. , With current estimate , ;

[0050] (1)

[0051] (2)

[0052] In the formula, , These represent the d-axis and q-axis currents, respectively. , These represent the voltages along the d-axis and q-axis, respectively. Indicates stator inductance; Indicates stator resistance; Indicates the magnetic flux linkage of the rotor permanent magnet; Indicates electric angular velocity; This represents the estimated electric angular velocity. , These represent the estimated current values ​​for the d-axis and q-axis, respectively. The differential variable is represented by the following parameters of the magnetic synchronous motor in this embodiment, as shown in Table 1 below.

[0053] Step 2: Subtract the adjustable model from the reference model to construct a nonlinear time-varying error feedback system:

[0054] (3)

[0055] In the formula, For current error, A= W is the nonlinear feedback channel matrix, and C is the forward channel gain matrix. The output signal of the forward channel is obtained by weighting and synthesizing the stator current estimation error vector e with the forward channel gain matrix C.

[0056] Then, the sliding surface is selected using Popov's hyperstability theorem:

[0057] (4)

[0058] Differentiate with respect to the sliding surface:

[0059] (5)

[0060] when At this point, the system enters the sliding mode and remains on the sliding surface. As defined by the sliding surface, the adjustable model of the motor tends to the reference model, and the electric angular velocity of the motor rotor tends to the actual electric angular velocity, thus achieving accurate observation of the rotor speed and position. Based on the selected sliding surface and the basic form of super-helical sliding mode control, a speed adaptive law is designed, and a control function that continuously switches at zero is employed. Replace the switching control function that is discontinuous at the zero point. The center frequency of SOGI can be obtained as follows:

[0061] (6)

[0062] In the formula, k1 and k2 are the superspiral sliding mode control gains.

[0063] The frequency response expression of the frequency-locked loop (FLL) in a traditional second-order generalized integral flux linkage observer is as follows:

[0064] (9)

[0065] In the formula, Let Γ be the estimated frequency of FLL, and Γ be the gamma function.

[0066] As can be seen from equation (9), the frequency response expression of the frequency-locked loop (FLL) contains a second harmonic oscillation term, which causes second harmonic fluctuations in the center frequency, resulting in the inability to accurately track the actual fundamental frequency of the input signal. In severe cases, it can cause the SOGI to deviate from the resonant operating point, thereby causing orthogonal signal extraction errors, position estimation deviations, and a decrease in system control performance, resulting in a loss of lock. In contrast to equation (6), this invention uses model reference adaptation to provide the center frequency for the SOGI. The frequency adaptation law does not contain a second harmonic oscillation term, so it can effectively suppress the fluctuation of the center frequency, reduce frequency tracking jitter, and thus improve the stability of the SOGI center frequency and its anti-loss-of-lock capability.

[0067] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A sensorless control method for a model-reference adaptive permanent magnet synchronous motor, characterized in that, Includes the following steps: S1: Based on the current equations in the synchronous rotating coordinate system, establish the reference model and the adjustable model to obtain the stator current. , and its estimated value , ; S2: According to , , , A nonlinear time-varying error feedback system is constructed, and the rotor flux electric angular velocity is designed using Popov's superstability theorem and the super-helical sliding mode control method. The adaptive law uses a sigmoid(s) function with smooth and continuous characteristics instead of the switching function in traditional sliding mode control. Reduce inherent chattering in sliding mode; S3: Calculation - Motor rotor electromotive force in coordinate system The input is then processed by SOGI, and the rotor flux electric angular velocity provided by S2 is used for integration. Using the center frequency of SOGI, it allows for the acquisition of accurate rotor flux estimation values. Calculate the real-time position and speed of the rotor of the permanent magnet synchronous motor; S4: Based on vector control technology, the estimated rotor flux calculated in S3. By performing flux orientation and using coordinate transformation, the stator current of the permanent magnet synchronous motor is decoupled and closed-loop controlled in a rotating coordinate system, thereby achieving sensorless control of the permanent magnet synchronous motor.

2. The sensorless control method for a model reference adaptive permanent magnet synchronous motor as described in claim 1, characterized in that: The specific method in S1 is to establish a reference model (1) with the actual electric angular velocity as a parameter and an adjustable model (2) with the estimated electric angular velocity as a parameter based on the stator current equation of the permanent magnet synchronous motor in the synchronous rotating coordinate system, as follows: (1) (2) In the formula, , These represent the stator currents along the d-axis and q-axis, respectively. , These represent the stator voltages along the d-axis and q-axis, respectively. Indicates stator inductance; Indicates stator resistance; Indicates the magnetic flux linkage of the rotor permanent magnet; This represents the electric angular velocity of the rotor flux linkage; , These represent the estimated stator current values ​​for the d-axis and q-axis, respectively. This represents an estimated value of the rotor flux linkage electric angular velocity; It represents the differential variable.

3. The sensorless control method for a model reference adaptive permanent magnet synchronous motor as described in claim 2, characterized in that: The method for constructing a nonlinear time-varying error feedback system in S2 is as follows: Subtract the adjustable model from the reference model to obtain equation (3): (3) In the formula, For the stator current estimation error, A= W is the nonlinear feedback channel matrix, and C is the forward channel gain matrix. The output signal of the forward channel is obtained by weighting and synthesizing the stator current estimation error vector e with the forward channel gain matrix C.

4. The sensorless control method for a model reference adaptive permanent magnet synchronous motor as described in claim 3, characterized in that: In S2, the sliding surface is selected based on Popov's hyperstability theorem: (4) Differentiating with respect to the sliding surface, we get: (5) when When the error feedback system enters the sliding mode and remains on the sliding surface, it can be seen from the definition of the sliding surface that the adjustable model of the motor tends to the reference model, and the electric angular velocity of the motor rotor tends to the actual electric angular velocity, thereby realizing accurate observation of the rotor speed and position.

5. The sensorless control method for a model reference adaptive permanent magnet synchronous motor as described in claim 4, characterized in that: S2 rotor flux electric angular velocity The formula for calculating the center frequency of SOGI is as follows: (6) In the formula, k1 and k2 are the superspiral sliding mode control gains.

6. The sensorless control method for a model reference adaptive permanent magnet synchronous motor as described in claim 1, characterized in that: S3 permanent magnet synchronous motor in - The equation for the rotor electromotive force of the motor in the coordinate system is: (7) In the formula, , They are permanent magnet synchronous motors axis, Shaft stator electromotive force; , They are respectively axis, Shaft stator current; , They are respectively axis, Shaft stator current.

7. The sensorless control method for a model reference adaptive permanent magnet synchronous motor as described in claim 6, characterized in that: The rotor flux estimate The calculation formula is as follows: (8) In the formula, , The rotor flux obtained by SOGI is respectively in The estimated value on the axis; k is the SOGI gain coefficient, used to adjust the dynamic response and filtering performance; s is the Laplace operator.

Images