Hybrid flux-based sensorless control method for permanent magnet synchronous motor in full speed range

Abstract

The application discloses a full-speed-range sensorless control method for a permanent magnet synchronous motor based on hybrid flux linkage. The method injects the electric angle information obtained by a square wave injection method into a current model of the flux linkage, replaces the original electric angle feedback, and forms a motor zero-speed flux linkage model based on the square wave injection method. Then, a PI control is used to add a voltage-current hybrid flux linkage model to constitute a full-speed-range flux linkage model to estimate rotor position information. Compared with a traditional linear weighting switching, the method is smoother, and a low-pass filter usually needed by a speed output of the traditional square wave injection method is omitted. The method can accurately estimate the rotor position information of the motor in a zero-low speed range, a medium-high speed range and a transition speed range, has good steady-state and dynamic performance, and can realize full-speed-range position sensorless control of the permanent magnet synchronous motor.

Description

Technical Field

[0001] This invention belongs to the field of permanent magnet synchronous motor drive control application, and in particular relates to a sensorless control method for permanent magnet synchronous motors in the full speed range based on hybrid flux linkage. Background Technology

[0002] Permanent magnet synchronous motors (PMSMs) possess advantages such as high efficiency, high torque density, and high dynamic performance. Driven by the need for carbon neutrality, their applications in transportation, aerospace, and other fields are becoming increasingly widespread. Replacing mechanical sensors with sensorless control technology has been a hot research topic in motor drive control. However, at zero-speed start-up or low speeds, the back EMF and flux linkage signal-to-noise ratios of PMSMs are low, making accurate observation difficult with back EMF and flux linkage observers. At medium to high speeds, signal injection methods are more complex than back EMF and flux linkage methods, and the injected signal further increases motor losses, making them unsuitable for medium to high speed ranges. To achieve full-speed-range operation of PMSMs with sensorless control, a composite control approach is needed, combining methods suitable for the zero-speed and low-speed ranges with those suitable for the medium-to-high-speed ranges.

[0003] Chinese invention patent CN119232009A, entitled "A Sensorless Control Method for Permanent Magnet Synchronous Motors in the Full Speed ​​Domain," discloses a sensorless control method for permanent magnet synchronous motors in the full speed domain. In the zero-low speed domain, a high-frequency pulse injection method is implemented, injecting high-frequency pulse signals into the motor system and extracting rotor position information based on the system response characteristics. In the medium-high speed domain, a super-spiral sliding mode observer method is used to estimate rotor position and speed in real time. A composite control strategy is employed to combine the two methods to achieve composite control in the full speed domain. Chinese invention patent CN119182323A, entitled "A Sensorless Control Method and System for Surface-Mounted Permanent Magnet Synchronous Motors in the Full Speed ​​Domain," discloses a sensorless control method and system for surface-mounted permanent magnet synchronous motors in the full speed domain. This method uses a high-frequency square wave voltage injection method for rotor position identification in the zero-low speed region, and an improved sliding mode observer for rotor position identification in the medium-high speed domain. In the switching speed domain, an improved weighted switching method with a continuously smooth proportional coefficient is used to achieve stable speed switching. This method has the advantage of accurately estimating motor speed and rotor position in the full speed domain. However, the aforementioned two weighted switching and other composite control methods inevitably suffer from problems such as insufficient stability in the transition zone between different observation methods.

[0004] Therefore, a new method is urgently needed to achieve sensorless control of permanent magnet synchronous motors across the entire speed range. Summary of the Invention

[0005] Purpose of the invention: In order to solve the problems existing in the prior art and realize sensorless control of permanent magnet synchronous motors in the full speed range, this invention proposes a sensorless control method for permanent magnet synchronous motors in the full speed range based on hybrid flux linkage.

[0006] Technical solution:

[0007] This invention proposes a sensorless control method for a permanent magnet synchronous motor across the entire speed domain based on hybrid flux linkage, comprising:

[0008] Based on the voltage equation and flux linkage equation of the motor, a voltage model suitable for the high-speed region and a current model suitable for the low-speed region are established respectively. The flux linkage difference between the current model and the voltage model is fed back to the back electromotive force of the voltage model through a PI controller for correction.

[0009] Using the high-frequency square wave injection method, the calculated estimated angle is input into the current model, and Park and inverse Park transformations are performed to replace the feedback of the originally estimated rotor position angle in the current model, thus constructing a square wave flux linkage model.

[0010] By using the voltage model as the output and the current model and square wave flux linkage model as compensation links, flux linkage correction is performed on the voltage model to construct a hybrid flux linkage model suitable for the entire speed domain.

[0011] The position angle and rotational speed of the hybrid flux model are extracted using a phase-locked loop, and the reference voltage vector is calculated. After processing, the driving voltage of the permanent magnet synchronous motor is obtained to control its operation.

[0012] Furthermore, the voltage equations of the motor include the voltage equations under the dq axis and the voltage equations under the αβ axis:

[0013] The equation for the voltage across the dq axis of a permanent magnet synchronous motor is expressed as follows:

[0014]

[0015] In the formula, u d u q Represents the dq axis voltage, i d i q R represents the dq-axis current. s L represents the stator resistance. q L represents the q-axis inductance. d This represents the d-axis inductance, and p represents the number of pole pairs of the motor. Indicates the electric angular velocity of the rotor. Indicates permanent magnet flux linkage;

[0016] The voltage equations under the αβ axis are obtained by inverse Park transform and simplification:

[0017]

[0018] In the formula, u α u β Represents the voltage along the α and β axes, i α i β This represents the αβ axis current.

[0019] Furthermore, establishing a voltage model suitable for the high-speed region includes:

[0020] The formula for stator flux linkage is:

[0021]

[0022] In the formula, , The stator flux linkage on the αβ axis, and the extended flux linkage below the αβ axis. , Relationship with stator flux linkage:

[0023] .

[0024] Furthermore, the current model suitable for the low-speed region is constructed using a coordinate transformation method, with the stator flux linkage along the dq axis. , The formula is:

[0025]

[0026] The flux linkage equation under the dq axis is obtained by combining the dq axis current and the fixed parameters of the motor. The flux linkage equation under the αβ axis is obtained by performing an inverse Park transform on the dq axis flux linkage equation. The matrix expression of the inverse Park transform is as follows:

[0027] .

[0028] Furthermore, the step of correcting the difference in flux linkage between the current model and the voltage model by feeding it back to the back electromotive force of the voltage model through a PI controller includes:

[0029] Once the motor starts running and reaches a steady state in the high-speed range, the hybrid flux linkage model of the motor can be expressed as the weighted sum of the current and voltage models, as follows:

[0030]

[0031] In the formula, k p With k i These are the coefficients of the PI controller. For stator flux linkage, The stator flux linkage is calculated using the current model. The stator flux linkage is calculated from the voltage model, and s is a complex variable in the Laplace transform. Under the regulation of the PI controller, the total flux linkage is regarded as a second-order combined flux linkage observer, and the real parts of the poles of the second-order filter are all negative.

[0032] Furthermore, the construction of the square wave flux linkage model includes:

[0033] The mathematical model for a motor in a zero-speed start-up state is expressed as follows:

[0034]

[0035] In the formula, u dh with u qh i is the high-frequency voltage of the motor on the dq axis. dh with i qh L represents the high-frequency current of the motor on the dq axis. dh For the d-axis inductance of the high-frequency square wave injection model, L qh The q-axis inductance of the high-frequency square wave injection model, and L dh =L d L qh =L q ;

[0036] When extracting the rotor information of the motor, the current equation on the dq axis is transformed to the αβ axis. Combining this with the known angular error between the estimated and actual positions, we obtain:

[0037]

[0038] In the formula, i represents the electrical angle error between the actual coordinate axes and the estimated coordinate axes. αh with i βh This refers to the high-frequency current of the motor along the αβ axis. and It estimates the voltage on the coordinate axes;

[0039] The square wave signal injected into the d-axis of the motor is represented as:

[0040]

[0041] In the formula, n is the sampling sequence number, the sampling frequency is twice the switching frequency, and the frequency of the injected square wave signal is the same as the switching frequency of the inverter, which simplifies to the square wave flux linkage model:

[0042]

[0043] In the formula, k is the sampling sequence number.

[0044] Furthermore, the construction of the hybrid flux linkage model applicable to the full speed domain includes:

[0045] Using the voltage flux linkage model as the output of the hybrid flux linkage model, and the current flux linkage model and the square wave flux linkage model as compensation components, flux linkage correction is performed on the voltage model. The overall flux linkage expression at this point is:

[0046]

[0047] In the formula, ψ sh For the square wave flux linkage model, the above equation divides the second-order system under PI feedback control into three modules, which, when converted to a time-domain model, yields:

[0048] .

[0049] Furthermore, the extraction of the position angle and rotational speed of the hybrid flux linkage model using a phase-locked loop includes:

[0050] The quadrature phase-locked loop includes a phase detector, a loop filter, and a voltage-controlled oscillator. The phase-locked loop automatically synchronizes the output signal with the input signal through closed-loop regulation, obtains the estimated speed by tracking the error through a PI controller, and obtains the estimated rotor position angle by integrating the estimated speed.

[0051] Furthermore, obtaining the drive voltage of the permanent magnet synchronous motor includes:

[0052] The estimated position angle is input into the transformation matrix to complete the conversion between the dq axis and the αβ axis. The calculated reference voltage vector is used as the SVPWM input to generate a PWM wave, which is then output as a three-phase voltage by a voltage source inverter to drive the motor and achieve sensorless control operation.

[0053] This invention also proposes a sensorless control system for a permanent magnet synchronous motor across the entire speed domain based on hybrid flux linkage, comprising:

[0054] The voltage and current flux linkage module is used to establish a voltage model suitable for the high-speed region and a current model suitable for the low-speed region based on the voltage equation and flux linkage equation of the motor, respectively. The flux linkage difference between the current model and the voltage model is fed back to the back electromotive force of the voltage model through a PI controller for correction.

[0055] The square wave flux module is used to input the calculated estimated angle into the current model using the high-frequency square wave injection method, perform Park and inverse Park transformations, replace the feedback of the original estimated rotor position angle in the current model, and construct the square wave flux model.

[0056] The hybrid module is used to take the voltage model as the output and the current model and square wave flux linkage model as compensation links to perform flux linkage correction on the voltage model and construct a hybrid flux linkage model suitable for the entire speed domain.

[0057] The control module is used to extract the position angle and rotational speed of the hybrid flux model using a phase-locked loop, calculate the reference voltage vector, and obtain the drive voltage of the permanent magnet synchronous motor to control its operation after processing.

[0058] Beneficial Effects: Compared with existing technologies, this invention combines a high-frequency signal injection method suitable for the zero-speed and low-speed domain with a flux linkage model observation method suitable for the medium-speed and high-speed domains. Specifically, it constructs an improved flux linkage observer by feeding back the flux linkage model obtained based on the square wave injection method to the voltage model, eliminating the need for additional switching control components. Compared to the traditional weighted average switching strategy, this invention constructs a hybrid flux linkage model across the entire speed domain, exhibiting excellent position estimation performance at different speeds and resolving the problem of uneven switching in the transition region. Attached Figure Description

[0059] Figure 1 A general block diagram for sensorless control of a permanent magnet synchronous motor across the entire speed range;

[0060] Figure 2 Block diagram of a hybrid flux observer;

[0061] Figure 3 Block diagram of a square wave flux observer;

[0062] Figure 4 Block diagram of a full-velocity hybrid flux observer;

[0063] Figure 5 The estimated stator flux linkage waveform is shown below.

[0064] Figure 6 The waveforms show the speed comparison from zero speed to rated speed;

[0065] Figure 7 A comparison chart of the actual and estimated position angles of the motor;

[0066] Figure 8 This is a diagram showing the motor position error. Detailed Implementation

[0067] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments.

[0068] like Figure 1 The diagram shown illustrates the sensorless full-speed-domain control of a permanent magnet synchronous motor based on hybrid flux, as proposed in this invention. The specific implementation steps of the control method proposed in this invention include:

[0069] Step 1: Construction of the voltage-current hybrid flux linkage model

[0070] The direct and quadrature axis inductances of the built-in permanent magnet synchronous motor are different, resulting in coupling terms in the voltage equation in the dq axis coordinate system. The voltage equation is as follows:

[0071]

[0072] In the formula, u d u q Represents the dq axis voltage, i d i q R represents the dq-axis current. s L represents the stator resistance. q L represents the q-axis inductance. d This represents the d-axis inductance, and p represents the number of pole pairs of the motor. Indicates the electric angular velocity of the rotor. This indicates the magnetic flux linkage of a permanent magnet.

[0073] As can be seen from the above equation, there are coupling terms in the voltage equation. In order to achieve decoupling, the above equation can be rewritten as shown in the following form:

[0074]

[0075] At this point, only the inductance parameter Lq remains in the resistance and inductance matrix, satisfying the symmetry principle. Based on this, the voltage equations along the αβ axis are obtained through the inverse Park transformation:

[0076]

[0077] In the formula, the second term on the right-hand side contains the derivative of the DC current. Since the current loop in the motor control responds rapidly to the current, it is assumed that the stator current is in a steady state, so this term is ignored, and the simplified formula is as follows:

[0078]

[0079] In the formula, u α u β Represents the voltage under the αβ axis, i α i β Represents the current along the αβ axis, θ e This represents the current electrical angle of the rotor. The rotor position information is extracted from the extended rotor flux linkage; the rotor flux linkage expression is:

[0080]

[0081] The voltage model is the most widely used model among motor flux linkage models. The stator flux linkage is obtained by integrating the back EMF of the motor. The calculation formula is as follows:

[0082]

[0083] In the formula, , Let be the stator flux linkage along the αβ axis. Then, based on the voltage equations mentioned earlier, we can obtain the extended flux linkage along the αβ axis. , Relationship with stator flux linkage:

[0084]

[0085] Unlike the voltage model, the current model constructs the flux linkage model through coordinate transformation. The specific stator flux linkage formula under the dq axis is as follows:

[0086]

[0087] By obtaining the dq-axis current and combining it with the motor's fixed parameters, the flux linkage equation for the dq-axis can be derived. Then, the flux linkage equation for the αβ-axis can be obtained through the inverse Park transformation. The inverse Park matrix expression is:

[0088]

[0089] Since the stator flux linkage is calculated, it needs to be converted into an extended flux linkage, similar to the voltage model, to obtain the rotor position angle. Unlike the voltage model, the current model uses the motor's inductance and permanent magnet flux linkage parameters in calculating the stator flux linkage, and requires Park and inverse Park transformations, making parameter errors more significant. The current model introduces feedback on the electrical angle, thus extending the lower limit of the motor's operating speed range compared to the voltage model. Furthermore, because no integral calculation is performed, it avoids the DC saturation problem of the voltage model. Compared to the voltage model, the current model does not involve stator resistance in its calculations but adds inductance and permanent magnet flux linkage. When operating in the high-speed domain, its anti-interference capability is insufficient, and the convergence problem caused by electrical angle feedback needs to be considered.

[0090] By combining the voltage and current models of magnetic flux, the advantages of both models can be fully utilized. The current model is used at low speeds, while the voltage model is switched at high speeds.

[0091] Once the motor starts running and reaches steady state, it operates in the high-speed range. At this point, the voltage model plays a primary role; therefore, the current model is chosen to correct the voltage model. The flux linkage difference between the current and voltage models is fed back to the back electromotive force of the voltage model via a PI controller for correction. The specific expression for the stator flux linkage is as follows:

[0092]

[0093] In the formula, k p With k i These are the coefficients of the PI controller. For stator flux linkage, The stator flux linkage is calculated using the current model. The stator flux linkage is calculated using the voltage model. To further analyze the hybrid flux linkage model, the above equation is transformed into the complex frequency domain, expressed as the weighted sum of the current and voltage models. The expression for the total flux linkage is as follows:

[0094]

[0095] In the formula, s is a complex variable in the Laplace transform, multiplied by a low-pass filter before the current model and by a high-pass filter before the voltage model. Therefore, under the regulation of the PI controller, the total flux linkage can be considered as a second-order combined flux linkage observer. Furthermore, to ensure system stability, the real parts of the poles of the second-order filter should all be negative.

[0096] If k i If set to 0, the control system structure is simplified to a first-order system, represented as:

[0097]

[0098] The block diagram of the hybrid flux observer is as follows: Figure 2 As shown, this hybrid model achieves closed-loop control of the flux linkage, effectively improving the motor's performance when transitioning from low-speed to high-speed operation.

[0099] Step 2: Construction of a square wave flux linkage model based on the square wave injection method

[0100] The hybrid flux observer overcomes the DC bias problem of the voltage flux observer and achieves good operation by introducing a current flux observer to correct the voltage flux observer. However, the current flux observer cannot achieve zero-speed start of the motor. The sensorless control of the motor in the full speed range is achieved by introducing a high-frequency square wave signal injection method.

[0101] When the motor is in a zero-speed start-up state, its speed is zero, and the back electromotive force generated by the motor speed can be ignored. The voltage drop across the inductor in the stator voltage is dominant, while the other voltage drops are negligible. Therefore, the mathematical model of the motor can be simplified as follows:

[0102]

[0103] After transformation, the high-frequency response current is obtained, and its expression is:

[0104]

[0105] In the formula, u dh with u qh i is the high-frequency voltage of the motor on the dq axis. dh with i qh L represents the high-frequency current of the motor on the dq axis. dhFor the d-axis inductance of the high-frequency square wave injection model, L qh The q-axis inductance of the high-frequency square wave injection model is L by default. dh =L d L qh =L q When extracting rotor information from the motor, calculations can be performed on either the dq-axis or the αβ-axis. However, calculations on the dq-axis require the use of estimated electrical angles, which reduces the iteration speed and can cause convergence problems. Therefore, this invention chooses the αβ-axis, which allows for direct calculation. The current equations on the dq-axis are transformed to the αβ-axis as shown in the following equation:

[0106]

[0107] In the formula, U dh U represents the d-axis stator voltage of the high-frequency square wave injection model. qh The above equation represents the q-axis stator voltage of the high-frequency square wave injection model. Given the angular error between the estimated position and the actual position, the above equation can be transformed into:

[0108]

[0109] In the formula, i represents the electrical angle error between the actual coordinate axes and the estimated coordinate axes. αh with i βh This refers to the high-frequency current of the motor along the αβ axis. and This estimates the voltage along the coordinate axes. The square wave signal injected into the motor's d-axis is represented as:

[0110]

[0111] In the formula, n is the sampling sequence number, the sampling frequency is twice the switching frequency, and the frequency of the injected square wave signal is the same as the switching frequency of the inverter. Substituting and simplifying the above two formulas, we obtain the square wave flux linkage model:

[0112]

[0113] In the formula, k is the sampling sequence number.

[0114] The error is obtained by decoupling through vector cross product, i.e.:

[0115]

[0116] In the formula, the right side is not zero when the motor has salient polarity. When the rotor angle error is small enough, it can be approximated as a function whose magnitude is proportional to the angle error. The phase-locked loop can be used to achieve tracking control of the electrical angle.

[0117] Based on the high-frequency square wave injection method, the estimated angle is input into the current model for Park and inverse Park transformations, replacing the feedback of the rotor position angle originally estimated by the current model. The control block diagram of the square wave flux linkage observer is shown below. Figure 3 As shown. The advantage of this flux linkage model is that it can perform addition and subtraction operations with current model flux linkage observers and voltage model flux linkage observers, thus extending the control range of the hybrid flux linkage model to the full speed domain.

[0118] Step 3, Construction of magnetic flux fusion model and sensorless control

[0119] Based on the aforementioned square wave flux linkage model, current flux linkage model, and voltage flux linkage model, the flux linkage observer is divided into three stages according to its application range: zero speed, low speed, and high speed. Since the motor will eventually stabilize in the high-speed domain, the voltage flux linkage model is used as the output of the flux linkage model, and the current flux linkage model and square wave flux linkage model are used as compensation links to perform flux linkage correction on the voltage model. The overall flux linkage expression at this time is as follows:

[0120]

[0121] In the formula, ψ sh For the square wave flux linkage model, the above equation divides the second-order system under PI feedback control into three modules, which can be converted into a time-domain model as follows:

[0122]

[0123] In the formula, k1 represents the compensation coefficient of the current model, and k2 represents the compensation coefficient of the high-frequency square wave injection model. The high-pass filter before the voltage model remains unchanged, while the low-pass filter before the original current model is split into a low-pass filter and a band-pass filter. The square wave flux linkage model is added to the total flux linkage after passing through the low-pass filter, and plays a major role when the motor is running in the zero-speed domain. After adding the square wave flux linkage model, the complexity of the hybrid flux linkage does not increase, and compared with directly using the square wave injection method, it eliminates the need for a low-pass filter to estimate the speed, thus reducing phase loss.

[0124] This model uses a square wave flux linkage model obtained through square wave injection during zero-speed startup for control. Although this increases the error caused by motor parameters compared to directly using the square wave injection method to estimate the electrical angle, the second-order low-pass filter also reduces harmonics during operation. Based on the motor's final stable speed, the voltage model is used as the output, and it is corrected by the current model and the zero-speed model based on the square wave injection method, ultimately obtaining a hybrid flux linkage model suitable for the entire speed range. Figure 4 As shown.

[0125] The extended flux linkage of the motor obtained through the hybrid flux linkage model inherently contains rotor position information. The position angle and speed are extracted using a quadrature phase-locked loop (quadrature phase-locked loop). The quadrature phase-locked loop includes a phase detector, loop filter, and voltage-controlled oscillator (VCO). It automatically synchronizes the output signal with the input signal through closed-loop regulation, obtains the estimated speed by tracking the error using a PI controller, and integrates the estimated speed to obtain the estimated rotor position. The position error obtained after normalizing the quadrature phase-locked loop is expressed as:

[0126]

[0127] In the formula, , The extended flux linkages along the α and β axes are respectively, Δθ e This represents the position error. The estimated motor speed is compared with the given speed, and then the current setpoint is obtained via a proportional-integral controller. Simultaneously, the estimated position angle is input into the transformation matrix to complete the conversion between the dq-axis and the αβ-axis.

[0128] The calculated reference voltage vector is used as the SVPWM input to generate a PWM wave, which is then output by a voltage source inverter to drive the motor, ultimately achieving sensorless control operation.

[0129] Figure 5 The obtained stator flux linkage waveform is estimated. As can be seen from the figure, the flux linkage waveform has good sinusoidal properties and there is no DC bias problem. Figure 6 The waveforms show the speed comparison from zero speed to rated speed. As can be seen from the figure, the estimated speed of the motor can follow the given speed well and remain stable at the rated speed.

[0130] Figure 7 and Figure 8 The figures show a comparison of the actual and estimated position angles of the motor, as well as a position error diagram. As can be seen from the figures, the estimated and actual position angles of the motor in the method proposed in this invention are basically coincident, with a maximum position error of approximately 0.2 rad, demonstrating good position estimation performance.

Claims

1. A sensorless control method for a permanent magnet synchronous motor across the entire speed domain based on hybrid flux linkage, characterized in that, include: Based on the voltage equation of the motor, a voltage model suitable for the high-speed region and a current model suitable for the low-speed region are established. The flux difference between the current model and the voltage model is fed back to the back electromotive force of the voltage model through a PI controller for correction. Using a high-frequency square wave injection method, the calculated estimated angle is input into the current model, and Park and inverse Park transformations are performed to replace the feedback of the originally estimated rotor position angle in the current model, thus constructing a square wave flux linkage model. The process of constructing the square wave flux linkage model includes: The mathematical model for a motor in a zero-speed start-up state is expressed as follows: In the formula, u dh with u qh i is the high-frequency voltage of the motor on the dq axis. dh with i qh L represents the high-frequency current of the motor on the dq axis. dh For the d-axis inductance of the high-frequency square wave injection model, L qh The q-axis inductance of the high-frequency square wave injection model, and L dh =L d L qh =L q L q L represents the q-axis inductance. d Indicates the d-axis inductance; When extracting the rotor information of the motor, the current equation on the dq axis is transformed to the αβ axis. Combining this with the known angular error between the estimated and actual positions, we obtain: In the formula, i represents the electrical angle error between the actual coordinate axes and the estimated coordinate axes. αh with i βh This refers to the high-frequency current of the motor along the αβ axis. and It estimates the voltage on the coordinate axes; The square wave signal injected into the d-axis of the motor is represented as: In the formula, n is the sampling sequence number, the sampling frequency is twice the switching frequency, and the frequency of the injected square wave signal is the same as the switching frequency of the inverter, which simplifies to the square wave flux linkage model: In the formula, k is the sampling sequence number, U in Input voltage; Using the voltage model as the output and the current model and the square wave flux linkage model as compensation links, flux linkage correction is performed on the voltage model to construct a hybrid flux linkage model suitable for the entire speed domain. The position angle and rotational speed of the hybrid flux model are extracted using a phase-locked loop, and the reference voltage vector is calculated. After processing, the driving voltage of the permanent magnet synchronous motor is obtained to control its operation.

2. The control method according to claim 1, characterized in that, The voltage equations of the motor include the voltage equations under the dq axis and the voltage equations under the αβ axis: The equation for the voltage across the dq axis of a permanent magnet synchronous motor is expressed as follows: In the formula, u d u q Represents the dq axis voltage, i d i q R represents the dq-axis current. s L represents the stator resistance. q L represents the q-axis inductance. d This represents the d-axis inductance, and p represents the number of pole pairs of the motor. Indicates the electric angular velocity of the rotor. Indicates permanent magnet flux linkage; The voltage equations under the αβ axis are obtained by inverse Park transform and simplification: In the formula, u α u β Represents the voltage along the α and β axes, i α i β This represents the αβ axis current.

3. The control method according to claim 2, characterized in that, The establishment of a voltage model suitable for the high-speed region includes: The formula for stator flux linkage is: In the formula, , The stator flux linkage on the αβ axis, and the extended flux linkage below the αβ axis. , Relationship with stator flux linkage: 。 4. The control method according to claim 3, characterized in that, The current model suitable for the low-speed region is constructed through coordinate transformation, with stator flux linkage under the dq axis. , The formula is: The flux linkage equation under the dq axis is obtained by combining the dq axis current and the fixed parameters of the motor. The flux linkage equation under the αβ axis is obtained by performing an inverse Park transform on the flux linkage equation under the dq axis. The matrix expression of the inverse Park transform is as follows: 。 5. The control method according to claim 4, characterized in that, The step of correcting the flux linkage difference between the current model and the voltage model by feeding it back to the back electromotive force of the voltage model via a PI controller includes: Once the motor starts running and reaches a steady state in the high-speed range, the hybrid flux linkage model of the motor can be expressed as the weighted sum of the current and voltage models, as follows: In the formula, k p With k i These are the coefficients of the PI controller. For stator flux linkage, The stator flux linkage is calculated using the current model. The stator flux linkage is calculated from the voltage model, and s is a complex variable in the Laplace transform. Under the regulation of the PI controller, the total flux linkage is regarded as a second-order combined flux linkage observer, and the real parts of the poles of the second-order filter are all negative.

6. The control method according to claim 5, characterized in that, The construction of the hybrid flux linkage model applicable to the full velocity domain includes: Using the voltage flux linkage model as the output of the hybrid flux linkage model, and the current flux linkage model and the square wave flux linkage model as compensation components, flux linkage correction is performed on the voltage model. The overall flux linkage expression at this point is: In the formula, ψ sh For the square wave flux linkage model, the above equation divides the second-order system under PI feedback control into three modules, which, when converted to a time-domain model, yields: 。 7. The control method according to claim 6, characterized in that, The extraction of the position angle and rotational speed of the hybrid flux linkage model using a phase-locked loop includes: The quadrature phase-locked loop includes a phase detector, a loop filter, and a voltage-controlled oscillator. The phase-locked loop automatically synchronizes the output signal with the input signal through closed-loop regulation, obtains the estimated speed by tracking the error through a PI controller, and obtains the estimated rotor position angle by integrating the estimated speed.

8. The control method according to claim 7, characterized in that, The process of obtaining the drive voltage for the permanent magnet synchronous motor includes: The estimated position angle is input into the transformation matrix to complete the conversion between the dq axis and the αβ axis. The calculated reference voltage vector is used as the SVPWM input to generate a PWM wave, which is then output as a three-phase voltage by a voltage source inverter to drive the motor and achieve sensorless control operation.

9. A sensorless control system for a permanent magnet synchronous motor across the entire speed range based on hybrid flux linkage, characterized in that, include: The voltage and current flux linkage modules are used to establish voltage models suitable for high-speed regions and current models suitable for low-speed regions based on the voltage equation and flux linkage equation of the motor, respectively. The flux linkage difference between the current model and the voltage model is fed back to the back electromotive force of the voltage model through a PI controller for correction. The square wave flux linkage module is used to input the calculated estimated angle into the current model using a high-frequency square wave injection method, perform Park and inverse Park transformations, and replace the feedback of the originally estimated rotor position angle in the current model to construct the square wave flux linkage model; the process of constructing the square wave flux linkage model includes: The mathematical model for a motor in a zero-speed start-up state is expressed as follows: In the formula, u dh with u qh i is the high-frequency voltage of the motor on the dq axis. dh with i qh L represents the high-frequency current of the motor on the dq axis. dh For the d-axis inductance of the high-frequency square wave injection model, L qh The q-axis inductance of the high-frequency square wave injection model, and L dh =L d L qh =L q L q L represents the q-axis inductance. d Indicates the d-axis inductance; When extracting the rotor information of the motor, the current equation on the dq axis is transformed to the αβ axis. Combining this with the known angular error between the estimated and actual positions, we obtain: In the formula, i represents the electrical angle error between the actual coordinate axes and the estimated coordinate axes. αh with i βh This refers to the high-frequency current of the motor along the αβ axis. and It estimates the voltage on the coordinate axes; The square wave signal injected into the d-axis of the motor is represented as: In the formula, n is the sampling sequence number, the sampling frequency is twice the switching frequency, and the frequency of the injected square wave signal is the same as the switching frequency of the inverter, which simplifies to the square wave flux linkage model: In the formula, k is the sampling sequence number, U in Input voltage; The hybrid module is used to take the voltage model as the output and the current model and square wave flux linkage model as compensation links to perform flux linkage correction on the voltage model and construct a hybrid flux linkage model suitable for the entire speed domain. The control module is used to extract the position angle and rotational speed of the hybrid flux model using a phase-locked loop, calculate the reference voltage vector, and obtain the drive voltage of the permanent magnet synchronous motor to control its operation after processing.

Images