Sensorless control method of supercoiled adaptive filter sliding mode observer

Abstract

The application discloses a kind of supercoil self-adapting filter sliding mode observer's position sensorless control method, it is related to, it is related to permanent magnet motor control technical field, first, the sliding mode observer based on supercoil algorithm is executed, through the integral and proportional term operation in discrete time, directly generate a continuous smooth original back electromotive force;Second, adaptive digital filtering is executed, according to the real-time monitoring motor angular velocity change rate, a group of wide or narrow bandwidth parameters is dynamically selected to configure filter, for processing the aforementioned smooth signal;Third, the real-time position and speed of rotor are accurately calculated by performing arctangent operation and phase-locked loop processing to the filtered signal, to calculate and generate the final PWM signal of driving motor as control instruction.The application can actively avoid chattering, thereby improving the control accuracy and stability of motor system.

Description

Technical Field

[0001] This invention relates to the field of permanent magnet motor control technology, and more specifically to a sensorless control method for a superspiral adaptive filter sliding mode observer. Background Technology

[0002] In traditional permanent magnet synchronous motor (PMSM) control systems, position information is typically obtained through mechanical position sensors. However, the installation of mechanical position sensors not only increases system costs but also negatively impacts system reliability and control accuracy. The emergence of sensorless control strategies overcomes the shortcomings of position sensors, such as poor environmental adaptability, high installation and maintenance costs, and low reliability, thus expanding the applicability of PMSM control systems. Sensorless control strategies are widely used in high-performance applications such as robot joint module motors, where there are stringent constraints on actuator size, weight, and reliability.

[0003] Sliding mode observers are one of the most widely used sensorless control schemes for permanent magnet synchronous motors (PMSMs) due to their strong robustness to motor parameters and external load disturbances. However, the control law of traditional first-order sliding mode observers typically relies on discontinuous sign functions, which inherently leads to chattering. This high-frequency chattering component is directly superimposed on the estimated back EMF signal and ultimately manifests as parasitic high-frequency torque pulsations through the current control loop. In precision robotic operations, these torque harmonics are extremely detrimental, directly reducing the system's control accuracy, trajectory tracking fidelity, and static position holding stability.

[0004] To attenuate torque ripple, existing technologies typically employ a strategy of cascading filters at the observer output. Using a low-pass filter to filter the estimated back EMF introduces frequency-dependent phase lag, severely limiting the dynamic bandwidth of the closed-loop control system and potentially impairing system stability. Furthermore, existing technologies generally follow a passive "generation-suppression" design paradigm, assuming chattering is an inherent product of sliding mode observation and then suppressing it through external compensation. This paradigm fails to fundamentally solve the problem; its theoretical flaw lies in the inability to construct an inherently continuous sliding mode control law to actively avoid chattering.

[0005] Therefore, how to generate a smooth and continuous back EMF estimate from within the observer, actively avoid chattering, and thus improve the control accuracy and stability of the system is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0006] In view of this, the present invention provides a sensorless control method for a superspiral adaptive filter sliding mode observer, which is applicable to robot joint module motors. It solves the inherent control chattering caused by the discontinuous control law of traditional positionless control algorithms based on sliding mode observers, as well as the inherent defects of subsequent filtering links in balancing dynamic response and steady-state accuracy.

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

[0008] A sensorless control method for a superspiral adaptive filter sliding mode observer includes the following steps:

[0009] Step 1: Collect the three-phase current and three-phase voltage of the permanent magnet synchronous motor;

[0010] Step 2: Convert the three-phase current and three-phase voltage into two-phase current components in the stationary coordinate system using Clark coordinate transformation. , ) and two-phase voltage components ( , );

[0011] Step 3: Combine the two-phase current components ( , ) and two-phase voltage components ( , The input is fed into the superspiral sliding mode observer to obtain the original back electromotive force and update the current estimates of the α-axis and β-axis.

[0012] Step 4: Perform adaptive digital filtering based on the historically estimated motor angular velocity and the original back EMF to obtain the current fundamental back EMF of the α-axis and β-axis;

[0013] Step 5: Calculate the rotor's estimated electric angular velocity and estimated rotor position angle using a phase-locked loop based on the fundamental back electromotive force of the α-axis and β-axis, and generate motor control commands.

[0014] Preferably, step 3 is based on the two-phase current components ( , ) and two-phase voltage components ( , The back electromotive force of a permanent magnet synchronous motor is observed using a super-helical sliding mode observer, and the current estimate is updated. This includes the following steps:

[0015] Step 31: Based on the two-phase current components ( , Calculate the current estimation errors for the α-axis and β-axis;

[0016] In each control cycle k, the current estimation error for the current cycle k is calculated based on the actual two-phase current components of the current cycle k and the estimated current values ​​on the α and β axes obtained from the previous cycle k-1; the expression is:

[0017] ;

[0018] in, and These represent the current estimation errors of the current in the current period k along the α-axis and β-axis, respectively. and These represent the estimated current values ​​on the α-axis and β-axis calculated in the previous cycle k-1, respectively; and These represent the α-axis and β-axis current components of the current period k obtained from actual measurements, respectively.

[0019] Step 32: Estimate the error based on the current. and The internal state variables of the superspiral algorithm, i.e., the integral terms, are updated. This update process is a discrete integral, and its expression is:

[0020] ;

[0021] in, and These represent the superspiral algorithm integral terms for the α-axis and β-axis after the current period k is updated; and These represent the superspiral algorithm integral terms for the α-axis and β-axis after the previous k-1 update, respectively; The integral gain of the superspiral algorithm is a pre-defined positive constant; sgn() is the sign function.

[0022] Step 33: Calculate the original back electromotive force based on the current estimation error and the updated integral term;

[0023] The current estimation error and the updated integral term are combined to generate the original back electromotive force on the α and β axes. This calculation integrates proportional and integral terms, and its expression is as follows:

[0024] ;

[0025] in, and These represent the original back electromotive forces along the α-axis and β-axis of the current period k, respectively, output by the superspiral sliding mode observer; This represents the pre-defined proportional gain of the superspiral algorithm, which is a positive constant. and These represent the square roots of the absolute values ​​of the current estimation errors for the α-axis and β-axis, respectively.

[0026] Step 34: Update the current estimate for the current cycle based on the original back electromotive force;

[0027] Using the electrical equations of a permanent magnet synchronous motor and combining the original back electromotive force, the current along the α and β axes of the current period k is estimated and updated to obtain the estimated current value. This update process is based on a discretized motor model, and its expression is as follows:

[0028] ;

[0029] in, and These represent the updated current estimates on the α and β axes for the current period k, respectively. These values ​​will be stored and used for current estimation error calculation in the next period k+1; L represents the stator inductance of the permanent magnet synchronous motor; R represents the stator resistance of the permanent magnet synchronous motor. This indicates the sampling period of the permanent magnet synchronous motor control system; and These represent the voltage components of the α-axis and β-axis applied to the permanent magnet synchronous motor in the previous cycle k-1, respectively.

[0030] Preferably, step 4 employs an adaptive digital filtering method with an infinite impulse response (IIR) digital filter that has an adaptive cutoff bandwidth to perform adaptive digital filtering. This method precisely extracts the fundamental component, which is closely related to the rotor position and velocity information, from the original back electromotive force output by the super-spiral sliding mode observer, which contains high-frequency noise and switching harmonics, thus obtaining the fundamental back electromotive force. Specifically, this includes the following steps:

[0031] Step 41: Dynamically adjust the cutoff bandwidth of the filter based on the historically estimated electric angular velocity of the motor;

[0032] The cutoff bandwidth of the filter is dynamically adjusted based on the rate of change of the motor's electrical angular velocity, expressed as:

[0033] ;

[0034] ;

[0035] in, It represents the absolute value of the estimated change in the electric angular velocity of the motor between two consecutive cycles; This represents the cutoff bandwidth of the filter to be used in the current period k, i.e., the dynamically adjusted cutoff bandwidth. , These represent the estimated electric angular velocities of the motor based on historical periods k-1 and k-2, respectively. This indicates the preset threshold for judging the rate of change of motor speed in dynamic / steady-state conditions; and These represent the preset wide cutoff bandwidth for the dynamic process of the motor and the narrow cutoff bandwidth for the steady-state process of the motor, respectively.

[0036] Step 42: Calculate or find the filter coefficients based on the motor's electric angular velocity and cutoff bandwidth. , , , , ;

[0037] The electric angular velocity of the motor is estimated based on the period k-1. and cutoff bandwidth The coefficients of the filter used in the current period k are determined by online real-time calculation or by consulting a pre-stored look-up table (LUT). , , , , ;

[0038] Step 43: Calculate the fundamental back EMF by performing a difference equation based on the filter coefficients and the historical raw back EMF;

[0039] The original back electromotive force (EMF) is filtered using the filter coefficients. Then, by executing the difference equation, the fundamental back EMF along the α and β axes after filtering for the current period k is calculated. The expression is:

[0040] ;

[0041] in, and These represent the fundamental back electromotive forces along the α and β axes after filtering for the current period k, respectively. and These represent the original back electromotive forces of the current period k output by the superspiral sliding mode observer on the α and β axes, respectively. , These represent the original back electromotive forces of period k-1 on the α-axis and β-axis, respectively; , These represent the original back electromotive forces of period k⁻² on the α-axis and β-axis, respectively; , These represent the fundamental back electromotive forces along the α and β axes after filtering, respectively, with a period of k-1. , These represent the fundamental back electromotive forces along the α and β axes after filtering, respectively, for a period of k-2.

[0042] Preferably, the specific process of step 5, which calculates the estimated rotor electric angular velocity and estimated rotor position angle based on the fundamental back electromotive forces of the α-axis and β-axis, and generates motor control commands, is as follows:

[0043] Step 51: Generate a phase error signal Δe by estimating the rotor position angle based on the fundamental back electromotive force of the α-axis and β-axis in the current cycle and the rotor position angle of the previous cycle k-1.

[0044] Based on the α-axis fundamental wave back electromotive force of the current period k and β-axis fundamental wave back electromotive force The position angle is estimated by combining the rotor position from the previous cycle k-1. The phase error signal Δe is calculated using the following expression:

[0045] ;

[0046] Step 52: Normalize the phase error signal to generate the position angle error value;

[0047] The phase error signal Δe is normalized by dividing it by an estimated back electromotive force amplitude, k. e To eliminate the impact of motor speed variations on the phase-locked loop gain, the expression is:

[0048] ;

[0049] in, This represents the position angle error value of the current period k; This represents the phase error signal for the current period k; This represents the estimated magnitude of the back electromotive force in the current period k.

[0050] Step 53: Based on the position angle error value, the rotor's estimated electrical angular velocity is generated by proportional-integral (PI) regulation in the phase-locked loop (PLL).

[0051] The phase error signal, after normalization, is input to a preset PI controller in the PLL. The output of the PI controller is the estimated rotor electrical angular velocity for the current period k. The expression is:

[0052] ;

[0053] in, This represents the proportional gain of the PI controller; This represents the cumulative value of the integral term calculated using the superspiral algorithm integral terms based on the α-axis and β-axis of the previous period k-1; This represents the integral gain of the PI controller; This indicates the sampling period of the permanent magnet synchronous motor control system, which is also the control period.

[0054] Step 54: Update the estimated rotor position angle based on the estimated rotor electrical angular velocity;

[0055] Estimating the electrical angular velocity of the rotor Perform discrete-time numerical integration, and compare the result of the numerical integration with the estimated rotor position angle from the previous cycle k-1. Add them together to update the estimated rotor position angle for the current period k. The expression is:

[0056] ;

[0057] Step 55: Generate motor control commands based on the estimated electric angular velocity of the rotor and the updated estimated rotor position angle.

[0058] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a sensorless control method for a super-spiral adaptive filter sliding mode observer. First, a sliding mode observer based on a super-spiral algorithm is executed, which directly generates a continuous and smooth original back EMF estimation signal through integral and proportional term operations in discrete time, so as to avoid the control chattering problem of traditional methods from the root. Second, an adaptive proportional resonant filtering step is executed. This step dynamically selects a set of wide or narrow bandwidth parameters to configure the filter according to the real-time monitored motor angular velocity change rate, which is used to process the aforementioned smooth signal. Third, by performing arctangent operation and phase-locked loop (PLL) processing on the filtered signal, the real-time position and speed of the rotor are accurately calculated. Finally, the high-precision, hysteresis-free position and speed information is input into the motor control system to calculate and generate the final PWM signal for driving the motor. This invention is applicable to robot joint module motors. By combining streamlined chatter-free estimation with adaptive filtering, it significantly improves the dynamic performance and control accuracy of robot joint module motors. It solves the inherent control chatter caused by the discontinuous control law of traditional position-free control algorithms based on sliding diaphragm observers, as well as the inherent defects of subsequent filtering stages in balancing dynamic response and steady-state accuracy. Attached Figure Description

[0059] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0060] Figure 1 A schematic flowchart of the sensorless control method for the superhelical adaptive filter sliding mode observer provided by the present invention;

[0061] Figure 2The flowchart of the superspiral adaptive filter sliding mode observer algorithm provided by this invention is shown. Detailed Implementation

[0062] 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.

[0063] This invention discloses a sensorless control method for a superspiral adaptive filter sliding mode observer, the process of which is as follows: Figure 1 As shown. In this embodiment, the electrical parameters of the permanent magnet motor are as follows: inductance Ld=Lq=0.3mH, resistance R=0.5mH. The rotor flux is =0.083Wb, number of permanent magnet pairs p=3, rated speed 300rad / s, control cycle T s =0.0002s, motor moment of inertia J=0.0001, friction coefficient B=0.0022, the specific control method is as follows:

[0064] S1: Collect the three-phase current and three-phase voltage of the permanent magnet synchronous motor;

[0065] S2: Convert the three-phase current and three-phase voltage into two-phase current components and two-phase voltage components in a stationary coordinate system through Clark coordinate transformation;

[0066] S3: Input the two-phase current components and two-phase voltage components into the super-helical sliding mode observer to obtain the original back electromotive force and update the current estimates of the α-axis and β-axis.

[0067] S4: Perform adaptive digital filtering based on the historically estimated motor angular velocity and original back EMF to obtain the current fundamental back EMF of the α-axis and β-axis;

[0068] S5: Based on the fundamental back electromotive force of the α-axis and β-axis, the rotor's estimated electric angular velocity and estimated rotor position angle are calculated through a phase-locked loop, and dual closed-loop PI control is performed to generate motor control commands.

[0069] In one specific embodiment, such as Figure 2 As shown, the back electromotive force of the permanent magnet synchronous motor is observed using a super-helical sliding mode observer, and the current estimate is updated. The specific execution process is implemented by the observer module in the permanent magnet synchronous motor control unit. This process receives the two-phase current components in the stationary coordinate system of S2. , ) and two-phase voltage components ( , As input, its detailed execution process includes the following steps:

[0070] S31: Based on the two-phase current components ( , ) Calculate the current estimation error;

[0071] In each control cycle k, the current estimation error for the current cycle k is calculated based on the actual two-phase current components of the current cycle k and the estimated current values ​​on the α and β axes obtained from the previous cycle k-1; the expression is:

[0072] ;

[0073] in, and These represent the current estimation errors of the current in the current period k along the α-axis and β-axis, respectively. and These represent the estimated current values ​​on the α-axis and β-axis calculated in the previous cycle k-1, respectively; and These represent the α-axis and β-axis current components of the current period k obtained from actual measurements, respectively.

[0074] S32: Estimating error based on current and The internal state variables of the superspiral algorithm, i.e., the integral terms, are updated. This update process is a discrete integral, and its expression is:

[0075] ;

[0076] in, and These represent the superspiral algorithm (STA) integral terms for the α-axis and β-axis after the current period k is updated, respectively, used to eliminate chattering and ensure finite-time convergence; and These represent the superspiral algorithm integral terms for the α-axis and β-axis after the previous k-1 update, respectively; This represents the pre-defined integral gain of the superspiral algorithm, which is a positive number; in this embodiment, it is set to 4. sgn() is the sign function.

[0077] S33: Calculate the original back electromotive force based on the current estimation error and the integral term;

[0078] The current estimation error and the updated integral term are combined to generate the original back electromotive force on the α and β axes. This calculation integrates proportional and integral terms, and its expression is as follows:

[0079] ;

[0080] in, and These represent the original back electromotive forces along the α and β axes of the current period k, respectively, output by the superhelical sliding mode observer. This signal is smooth and free of chattering, but may still contain harmonics. This represents the pre-set proportional gain of the superspiral algorithm, which is a normal number; in this embodiment, the value is 500. and These represent the square roots of the absolute values ​​of the current estimation errors for the α-axis and β-axis, respectively.

[0081] S34: Update the current estimate for the current cycle based on the original back electromotive force;

[0082] Using the electrical equations of a permanent magnet synchronous motor and combining the original back electromotive force, the current along the α and β axes of the current period k is estimated and updated to obtain the estimated current value. This update process is based on a discretized motor model, and its expression is as follows:

[0083] ;

[0084] in, and These represent the updated current estimates on the α and β axes for the current period k, respectively. These values ​​will be stored and used for current estimation error calculation in the next period k+1; L represents the stator inductance of the permanent magnet synchronous motor, which is 0.3mH in this embodiment; R represents the stator resistance of the permanent magnet synchronous motor, which is 0.5mH in this embodiment. ; This represents the sampling period of the permanent magnet synchronous motor control system, which is 0.0002s in this embodiment; and These represent the voltage components of the α-axis and β-axis applied to the permanent magnet synchronous motor in the previous cycle k-1, respectively.

[0085] In one specific embodiment, such as Figure 2 As shown, S4 employs an infinite impulse response (IIR) digital filter with adaptive cutoff bandwidth to perform adaptive digital filtering. From the raw back electromotive force output by the super-spiral sliding mode observer, which contains high-frequency noise and switching harmonics, the fundamental component closely related to rotor position and velocity information is accurately extracted to obtain the fundamental back electromotive force. Specifically, the following steps are included:

[0086] S41: Dynamically adjust the cutoff bandwidth of the filter based on historically estimated motor angular velocity;

[0087] To ensure the filter has a fast response speed during motor dynamic processes (such as acceleration and deceleration) and strong noise suppression capability during steady-state operation, the filter's cutoff bandwidth is dynamically adjusted according to the rate of change of the motor's electrical angular velocity. The expression is as follows:

[0088] ;

[0089] ;

[0090] in, It represents the absolute value of the estimated change in the electric angular velocity of the motor between two consecutive cycles; This indicates the cutoff bandwidth of the filter to be used in the current period k, i.e., the dynamically adjusted cutoff bandwidth. In this embodiment... =50 rad / s; , These represent the estimated electric angular velocities of the motor based on historical periods k-1 and k-2, respectively. This indicates the preset threshold for judging the rate of change of motor speed in dynamic / steady-state conditions; and These represent the preset wide cutoff bandwidth for the dynamic process of the motor and the narrow cutoff bandwidth for the steady-state process of the motor, respectively, in this embodiment. =300 rad / s, =30 rad / s;

[0091] S42: Calculate or find the filter coefficients based on the motor's electric angular velocity and cutoff bandwidth. , , , , The electric angular velocity of the motor estimated based on the period k-1 and cutoff bandwidth The coefficients of the filter used in the current period k are determined by online real-time calculation or by consulting a pre-stored look-up table (LUT). , , , , The calculation uses a fourth-order Butterworth filter, and the coefficients are generated in MATLAB based on the input cutoff frequency and bandwidth.

[0092] S43: Calculate the fundamental back EMF by performing a difference equation based on the filter coefficients and the historical raw back EMF;

[0093] The original back electromotive force (EMF) is filtered using the filter coefficients. Then, by executing the difference equation, the fundamental back EMF along the α and β axes after filtering for the current period k is calculated. The expression is:

[0094] ;

[0095] in, and These represent the fundamental back electromotive forces along the α and β axes after filtering for the current period k, respectively. and These represent the original back electromotive forces of the current period k output by the superspiral sliding mode observer on the α and β axes, respectively. , These represent the original back electromotive forces of period k-1 on the α-axis and β-axis, respectively; , These represent the original back electromotive forces of period k⁻² on the α-axis and β-axis, respectively; , These represent the fundamental back electromotive forces along the α and β axes after filtering, respectively, with a period of k-1. , These represent the fundamental back electromotive forces along the α and β axes after filtering, respectively, for a period of k-2.

[0096] In one specific embodiment, such as Figure 2 As shown, the specific process in S5 for calculating the rotor's estimated electric angular velocity and estimated rotor position angle based on the fundamental back electromotive forces of the α and β axes, and generating motor control commands, is as follows:

[0097] S51: Generate a phase error signal Δe based on the fundamental back electromotive force of the α-axis and β-axis in the current cycle and the rotor position angle estimated in the previous cycle k-1.

[0098] Based on the α-axis fundamental wave back electromotive force of the current period k and β-axis fundamental wave back electromotive force The position angle is estimated by combining the rotor position from the previous cycle k-1. The phase error signal Δe is calculated using the following expression:

[0099] ;

[0100] The mathematical transformation relationship mentioned above physically corresponds to the projection of the fundamental back electromotive force in the stationary coordinate system onto the cross axis (q-axis) of the synchronous rotating coordinate system with the rotor's estimated position angle as the reference. The amplitude of the generated phase error signal Δe is proportional to the sine value of the deviation when the deviation is small, thus effectively linearizing the angle deviation into a measurable electrical quantity.

[0101] S52: Normalize the phase error signal to generate the position angle error value;

[0102] The phase error signal Δe is normalized by dividing it by an estimated back electromotive force amplitude, k. e To eliminate the impact of motor speed variations on the phase-locked loop gain, the expression is:

[0103] ;

[0104] in, This represents the position angle error value of the current period k; This represents the phase error signal for the current period k; This represents the estimated magnitude of the back electromotive force in the current period k.

[0105] S53: Based on the position angle error value, the rotor's estimated electrical angular velocity is generated by proportional-integral (PI) regulation in the phase-locked loop (PLL).

[0106] The phase error signal, after normalization, is input to a preset PI controller in the PLL. The output of the PI controller is the estimated rotor electrical angular velocity for the current period k. The expression is:

[0107] ;

[0108] in, This represents the proportional gain of the PI controller, which is set to 70 in this embodiment. This represents the cumulative value of the integral term calculated using the superspiral algorithm integral terms based on the α-axis and β-axis of the previous period k-1; This represents the integral gain of the PI controller, which is set to 2500 in this embodiment. This indicates the sampling period of the permanent magnet synchronous motor control system, which is also the control period.

[0109] S54: Update the estimated rotor position angle based on the estimated rotor electrical angular velocity;

[0110] Estimating the electrical angular velocity of the rotor Perform discrete-time numerical integration, and compare the result of the numerical integration with the estimated rotor position angle from the previous cycle k-1. Add them together to update the estimated rotor position angle for the current period k. The expression is:

[0111] ;

[0112] S55: Generate motor control commands based on the estimated electric angular velocity and estimated position angle of the rotor.

[0113] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0114] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A sensorless control method for a superspiral adaptive filter sliding mode observer, characterized in that, Includes the following steps: Step 1: Collect the three-phase current and three-phase voltage of the permanent magnet synchronous motor; Step 2: Convert the three-phase current and three-phase voltage into two-phase current components and two-phase voltage components in the stationary coordinate system through coordinate transformation; Step 3: Input the two-phase current components and two-phase voltage components into the super-helical sliding mode observer to obtain the original back electromotive force and update the current estimates of the α-axis and β-axis. Step 4: Perform adaptive digital filtering based on the historically estimated motor angular velocity and the original back EMF to obtain the current fundamental back EMF of the α-axis and β-axis; Step 5: Calculate the rotor's estimated electrical angular velocity and estimated rotor position angle using a phase-locked loop based on the fundamental back electromotive force of the α-axis and β-axis, and generate motor control commands; Step 4 employs an infinite impulse response digital filter with adaptive cutoff bandwidth to perform adaptive digital filtering. The specific process is as follows: Step 41: Dynamically adjust the cutoff bandwidth of the filter based on the historically estimated electric angular velocity of the motor; Step 42: Calculate or find the filter coefficients based on the motor's electric angular velocity and cutoff bandwidth; Step 43: Calculate the fundamental back EMFs of the α-axis and β-axis by performing difference equations based on the filter coefficients and the historical raw back EMFs.

2. The sensorless control method for the superspiral adaptive filter sliding mode observer according to claim 1, characterized in that, The specific process of step 3 is as follows: Step 31: Calculate the current estimation error along the α-axis and β-axis based on the two-phase current components; Step 32: Update the integral terms of the superhelical algorithm for the α-axis and β-axis in the superhelical sliding mode observer based on the current estimation error; Step 33: Calculate the original back electromotive force along the α-axis and β-axis based on the current estimation error and the updated integral term; Step 34: Update the current estimates for the α-axis and β-axis of the current cycle based on the original back electromotive force.

3. The sensorless control method for the superspiral adaptive filter sliding mode observer according to claim 1, characterized in that, The specific process of step 5 is as follows: Step 51: Generate a phase error signal by estimating the position angle based on the fundamental back electromotive force of the α-axis and β-axis in the current cycle and the rotor position angle in the previous cycle. Step 52: Normalize the phase error signal to generate the position angle error value; Step 53: Based on the position angle error value, the rotor's estimated electrical angular velocity is generated by proportional-integral regulation through the proportional-integral regulator in the phase-locked loop; Step 54: Update the estimated rotor position angle based on the estimated rotor electrical angular velocity; Step 55: Generate motor control commands based on the estimated electric angular velocity of the rotor and the updated estimated rotor position angle.

Images