A sensorless adaptive starting method for high-speed permanent magnet synchronous motors

Abstract

This invention discloses a sensorless adaptive starting method for a high-speed permanent magnet synchronous motor. The method first... I‑f A high-speed permanent magnet synchronous motor under starting conditions is modeled. Based on this, a speed differential damper is designed to compensate for the reference speed, increase the system damping ratio, and improve starting stability. Then, an extended state observer-phase-locked loop structure is constructed to demodulate the position, speed, and speed differential, providing a basis for the adaptive starter controller and the speed differential damper. Finally, an adaptive starter controller is designed to coordinate the adjustment of the current vector amplitude and acceleration, ensuring that the motor always operates near the maximum torque-to-current ratio operating point, balancing stability, speed, and high efficiency.

Description

Technical Field

[0001] This invention belongs to the field of motor system control technology, specifically relating to a sensorless adaptive starting method for a high-speed permanent magnet synchronous motor. Background Technology

[0002] High-speed permanent magnet synchronous motors (PMSMs) provide pressurized air to fuel cell air compressors and are the core power source for hydrogen fuel cell compressors. Their performance affects the overall energy conversion efficiency and operational quality of the fuel cell. On one hand, the high-speed PMSM and air compressor form a compressor system, with speeds typically exceeding 100 kr / min, making the use of mechanical position sensors difficult, thus necessitating sensorless control. On the other hand, as a core component of the air compressor, the high-speed PMSM needs to quickly transition to higher speeds after startup to increase compressor pressure. Especially for air compressors using air bearings, friction is inevitable during the process from bearing standstill to takeoff (speed reaching 20 kr / min and above), directly affecting bearing life. Therefore, researching the rapid and stable startup of high-speed PMSMs is of great significance.

[0003] Open-loop compressors are commonly used in low-cost applications such as air compressors and fans. If Control as an auxiliary starting method, traditional IfThe current-frequency starting method (reference [WANG Zihui, LU Kaiyuan, BlAABJERG F. Asimple startup strategy based on current regulation for back-EMF-based sensorless control of PMSM[J]. IEEE Transactions on Power Electronics, 2012,27(8): 3817-3823]) lacks closed-loop feedback regulation capability, is prone to loss of synchronization at the critical point due to insufficient load capacity, and cannot adapt to variable load conditions. To adapt to variable load conditions, some studies have attempted to adjust the current amplitude based on angle information (reference [NAIR SV, HATUA K, PRASAD N. A Quick If Starting of PMSM DriveWith Pole Slipping Prevention and Reduced Speed ​​Oscillations[J]. IEEE Transactions on Industrial Electronics, 2021, 68(8): 6650-6661]). However, because the transition phase and acceleration phase are independent of each other, the acceleration abruptly drops to 0 at the initial moment of entering the transition phase, causing speed overshoot, angle difference, and current spikes. Furthermore, due to traditional... If The controlled motor system has a small damping ratio, and speed oscillation is... If The inherent characteristics of control; some studies have made predictive compensation for speed (reference [Wang Meng, Yang Jiaqiang, Zhang Xiang. A closed-loop If control method for current vector of surface-mounted permanent magnet synchronous motor [J]. Proceedings of the CSEE, 2015, 35(10):2513-2521]), but there are problems such as difficulty in pre-adjusting coefficients and high noise.

[0004] Therefore, it can be seen that the current If When dealing with different load requirements, startup algorithms still struggle to balance stability, speed, and efficiency. Designing an adaptive startup method that can dynamically adjust current amplitude and acceleration to achieve efficient, smooth, and fast startup remains a challenge. Summary of the Invention

[0005] In view of the above, the present invention provides a sensorless adaptive starting method for high-speed permanent magnet synchronous motors, which utilizes an extended state observer (ESO)-phase-locked loop (PLL) structure to obtain the speed derivative while demodulating the rotor position and speed, and then uses the speed derivative for compensation. IfThe reference speed at startup suppresses speed oscillations and improves startup stability. Simultaneously, based on the corrected reference position and the angle difference between the rotor position demodulated by the extended state observer-phase-locked loop, the acceleration and current amplitude are dynamically adjusted, balancing stability, speed, and high efficiency. Furthermore, weights are adaptively allocated according to speed to improve the dynamic performance during the transition from acceleration to the transition phase. In addition, from a control theory perspective... If The motor system under starting conditions is modeled and analyzed, and a method for tuning the compensation coefficient of the acceleration damper is given to reduce the complexity of parameter adjustment.

[0006] A sensorless adaptive starting method for a high-speed permanent magnet synchronous motor includes the following steps:

[0007] (1) According to If A mathematical model of a high-speed permanent magnet synchronous motor is developed for starting. A speed differential damper is designed to correct the reference speed and increase the system damping ratio. The compensation coefficient of the damper is then calculated and determined. k ;

[0008] (2) Construct an extended state observer-phase-locked loop structure. Use this structure to demodulate rotor position and speed information while obtaining a smooth speed differential signal, providing feedback compensation basis for the speed differential damper.

[0009] (3) Real-time detection feedback after compensation If By referencing the angle difference between the virtual coordinate system and the rotor position orientation coordinate system, an adaptive starting controller is constructed. Based on MTPA (maximum torque-to-current ratio), the current vector magnitude and acceleration are adaptively adjusted, and the q-axis reference current is calculated and determined. i q_ref and reference rotor position angle i ref Used to control the motor inverter.

[0010] Furthermore, the expression for the rotational speed differential damper in step (1) is as follows:

[0011]

[0012] in: oh i The rotating electric angular velocity of the stator current vector. oh i0 The reference speed before correction. Motor speed oh r The first differential.

[0013] Furthermore, the compensation coefficient k The calculation expression is as follows:

[0014]

[0015] in: K T The torque coefficient of the motor and K T =1.5 P n ψ f , J Let be the moment of inertia of the motor. P n This represents the number of pole pairs of the motor. I s0 and i err0 These are the motors after they enter steady state. I s and i err , I s The magnitude of the stator current vector. i err for If The angle difference between the reference virtual coordinate system and the rotor position orientation coordinate system ψ f It is the permanent magnet flux linkage of the motor.

[0016] Furthermore, the expression for the extended state observer-phase-locked loop structure in step (2) is as follows:

[0017]

[0018] in: Rotor position angle i r The observed values, Motor speed oh r The observed values, For the observations of the extended state variables, , , They are respectively , , The first-order differential, β 1. β 2. β 3 represents the coefficients of the extended state observer and , e 1 represents the angle difference between the observed and actual rotor position angles. I s1 Given the initial value of the reference current, oh c To expand the bandwidth of the state observer.

[0019] Furthermore, the angle difference e The expression for calculating 1 is as follows:

[0020]

[0021] in: e α and e β These are the α-axis back EMF and β-axis back EMF estimated by the sliding mode observer, respectively.

[0022] Furthermore, in step (3), the feedback compensation is... If The angle difference between the reference virtual coordinate system and the rotor position orientation coordinate system ,in i i In order to pass through oh i The electric angle obtained by integration, oh i The rotating electric angular velocity of the stator current vector is calculated using a rotational differential damper, i.e. , oh i0 This is the reference speed before correction.

[0023] Furthermore, the reference rotational speed oh i0 The calculation expression is as follows:

[0024]

[0025]

[0026] in: α i This is the reference value for rotor angular acceleration. K p1 and K i1 Given the proportional coefficient and integral coefficient, b These are the weighting coefficients. t Indicates time.

[0027] Furthermore, the weighting coefficients b The expression is as follows:

[0028]

[0029] in: oh ref Given the target rotational speed.

[0030] Furthermore, in step (3), the q-axis reference current is determined. i q_refand reference rotor position angle i ref Specifically: when the observed motor speed Reach the target speed oh ref And when the motor enters a steady state, , Before this, , ; I s The magnitude of the stator current vector and , The closed-loop reference current amplitude is determined by... The difference is obtained after PI (proportional-integral) control. I comp This is the magnitude compensation amount for the stator current vector, i.e.:

[0031]

[0032] in: K p2 and K i2 These are the given proportional coefficient and integral coefficient, respectively.

[0033] Furthermore, the specific method for controlling the motor inverter in step (3) is as follows: first, the three-phase stator current of the motor is collected and converted into the α-axis stator current. i α and β-axis stator current i β ,make i ref As the coordinate transformation angle i α and i β Converted to d-axis stator current i d and q-axis stator current i q Then set the d-axis reference current. i d_ref =0, respectively for and The d-axis stator voltage is obtained by PI control of the current error. u d and q-axis stator voltage u q ; Utilizing coordinate transformation angles i ref Will u d and u q Converted to α-axis stator voltage uα and β-axis stator voltage u β Finally based on u α and u β A series of PWM (Pulse Width Modulation) signals are generated by SVPWM (Space Vector Pulse Width Modulation) to drive the power switching devices in the motor inverter.

[0034] A computer device includes a memory and a processor, wherein the memory stores a computer program and the processor executes the computer program to implement the above-described sensorless adaptive starting method for a high-speed permanent magnet synchronous motor.

[0035] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described sensorless adaptive starting method for a high-speed permanent magnet synchronous motor.

[0036] Based on the above technical solution, the present invention has the following beneficial technical effects:

[0037] 1. The ESO-PLL structure proposed in this invention can obtain a relatively smooth speed derivative while demodulating speed and position information. The demodulated speed derivative is then fed back to the reference speed via a damper, increasing system damping and suppressing speed oscillations. Furthermore, this invention addresses speed oscillations from a control theory perspective. If The motor system under starting conditions was modeled and analyzed, and a method for tuning the parameters of the speed differential damper compensation coefficient was given, which reduced the complexity of parameter adjustment.

[0038] 2. This invention adaptively adjusts the motor acceleration and current amplitude during the starting process, and simultaneously controls the motor to operate near the maximum torque-to-current ratio operating point throughout the starting process based on the weight of the speed distribution regulator. This balances stability, speed, and high efficiency, achieving a smooth transition from the acceleration phase to the constant speed phase and a seamless connection between the open-loop and closed-loop operation. Attached Figure Description

[0039] Figure 1 This is a control structure block diagram of the sensorless adaptive starting method for high-speed permanent magnet synchronous motors according to the present invention.

[0040] Figure 2 The experimental waveform diagram shows the high-speed permanent magnet synchronous motor starting up to 20000 r / min using the method of the present invention (a load that increases with the speed is applied at the same time as starting).

[0041] Figure 3 The method of this invention enables a high-speed permanent magnet synchronous motor to operate in a steady state at 20000 r / min from an open-loop state. IfExperimental waveform diagram of switching the control to dual closed-loop control (the load is maintained at 0.571 Nm before and after the switch). Detailed Implementation

[0042] To describe the present invention in more detail, the technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0043] Example:

[0044] like Figure 1 As shown, in this embodiment, a voltage vector is first applied to attract the rotor to a specific position; this stage is the positioning stage. During the acceleration stage, both switches are in position 1. i q_ref Equal to the given current vector magnitude I s The angle of coordinate transformation is the position angle of the dq coordinate system. i i .

[0045] The sensorless adaptive starting method for high-speed permanent magnet synchronous motors in this embodiment is implemented as follows:

[0046] (1) Establish If A mathematical model of a high-speed permanent magnet synchronous motor is used for starting. Based on this model, a speed differential damper is designed to correct the reference speed, thereby increasing the system damping ratio and improving starting stability.

[0047] If The mechanical system model of the high-speed permanent magnet synchronous motor under starting conditions is as follows:

[0048]

[0049] in: I s Stator current vector i s amplitude, oh i The rotating electric angular velocity of the stator current vector. i i The electrical angle is obtained by integrating the rotating electric angular velocity of the stator current vector. oh r This is the actual rotor speed. i r The rotor position angle, i err for i r and i i difference, J It is the moment of inertia. B It is the damping coefficient. TL It is the load torque. P n It is an extreme logarithm. K T =1.5 P n ψ f This is the torque coefficient.

[0050] If The linearized small-signal model of the mechanical system of the high-speed permanent magnet synchronous motor under starting conditions is as follows:

[0051]

[0052] in: I s0 and q err0 This represents the current amplitude and angle difference at the equilibrium point.

[0053] The speed differential damper is designed as follows:

[0054]

[0055] in: oh i0 The reference speed before correction. oh i This is the reference speed after correction via speed derivative feedback. k This is the compensation coefficient.

[0056] After introducing a speed differential damper If The small-signal model of the mechanical system of the high-speed permanent magnet synchronous motor under starting conditions is as follows:

[0057]

[0058] Write down the result after introducing the speed differential damper i err about I s , T L and oh i The expression is as follows:

[0059]

[0060] The characteristic equation of the system is:

[0061]

[0062] Based on the characteristic equation of a typical second-order system s 2 +2 outside n s + oh n 2 =0, where ω n For the system bandwidth, x Given the system's damping ratio, write down the damping ratio after correction using a speed differential damper. x c for:

[0063]

[0064] In this embodiment, the damping ratio is designed to be... Therefore, the compensation coefficient k The design is as follows:

[0065]

[0066] (2) Construct an extended state observer-phase-locked loop structure to replace the orthogonal phase-locked loop in the traditional sliding mode observer. While demodulating the rotor position and speed information, a smooth speed differential signal is obtained, providing feedback compensation basis for the speed differential damper.

[0067] The system expression before state expansion is as follows:

[0068]

[0069] in: T L For unknown external disturbances d ( t ), this To control the input, f ( i , oh , d ( t ), t The disturbance is a lumped disturbance, which is then expanded into a new state variable. f Its derivative is g ( t ).

[0070] The expression of the system after state expansion is:

[0071]

[0072] The input to the extended state observer-phase-locked loop is the back electromotive force estimated by the sliding mode observer. e α and e β First, the normalized angle difference is obtained using the heterodyne method. ,in For the observed rotor position angle, the expression for the heterodyne method is:

[0073]

[0074] The expression for the extended state observer is:

[0075]

[0076] in: and These are the observed values ​​of rotational speed and extended state variables, respectively. β 1. β 2. β 3 represents the coefficients of the third-order extended state observer mentioned above; based on the expression, write the characteristic equation and design the bandwidth of the first extended state observer. oh c Therefore, the coefficients for designing the observer are... .

[0077] The output of the extended state observer-phase-locked loop is the rotor position observation value. Rotational speed observation value and the differential observation of rotational speed Observations of the differential rotational speed It can provide feedback compensation for the speed differential damper.

[0078] (3) Real-time detection feedback after compensation If An adaptive start-up controller is constructed by referencing the difference angle between the virtual coordinate system and the rotor position-oriented coordinate system demodulated by the extended state observer-phase-locked loop. The current vector magnitude and acceleration are adaptively adjusted based on the maximum torque-current ratio.

[0079] The position angle of the coordinate system oriented by the rotor position is obtained from the output of the extended state observer-phase-locked loop. replace, If The angle difference between the reference virtual coordinate system and the coordinate system oriented by the rotor position . i err =0 is the critical stability point and also the operating point of maximum torque-to-current ratio, where efficiency is highest. Therefore, the control objective of the adaptive starter controller is set as the input. The controller input is an error signal. The controller consists of an acceleration regulator and a current vector amplitude regulator, with outputs of acceleration and current vector amplitude respectively. α i and current vector magnitude compensation I comp The adaptive starter controller is designed as follows:

[0080]

[0081] in: K p1 、K i1 、K p2 、K i2 This is the PI adjustment coefficient of the adaptive controller. b These are the weighting coefficients of the adaptive starter controller, and their expressions are as follows:

[0082]

[0083] in: oh ref The target rotational speed is set.

[0084] The reference rotational speed for obtaining the reference stator current vector oh i0 Amplitude I s and coordinate transformation angle i i as follows:

[0085]

[0086] in: I s1 It is the initial value of the current vector amplitude, which is usually set as the rated current.

[0087] Once the target speed is reached, the switch switches to... Figure 1 Position 2 in the figure uses SMO sensorless control, with a reference rotational speed of 1000 rpm. oh ref Observer estimates As feedback rotational speed, the position angle in the dq coordinate system i r (Observer estimates angle) (replace) is the coordinate transformation angle i ref ,at the same time i q_ref It is obtained from the output of the speed loop PI.

[0088] Verification example:

[0089] To verify the effectiveness and superiority of the control method of this invention, we conducted experimental verification. The parameters of the surface-mounted high-speed permanent magnet synchronous motor used as an example in the experiment are shown in Table 1:

[0090] Table 1

[0091]

[0092] The experimental waveform of the high-speed permanent magnet synchronous motor starting to 20000 r / min using the method of this invention is as follows: Figure 2 As shown, a three-phase power resistor is connected throughout the startup process to simulate the fan load characteristics that vary with speed (20 kr / min corresponds to a load torque of 0.571 Nm). Figure 2 From top to bottom, the waveforms are: rotational speed, coordinate system angle difference, and q-axis current waveform. It can be seen that the motor speed steadily increases under varying load conditions, and during acceleration, the motor always operates near the maximum torque-to-current ratio operating point. Acceleration begins to decrease at 0.55s and reaches zero at 0.67s. The entire process is smooth, with virtually no overshoot at the moment of reaching the target speed. The steady-state time to 20000 r / min is approximately 0.71s. The steady-state amplitude of the phase current is 26A, a 35% reduction compared to the initial value. The steady-state angle difference is 0.16π rad. Throughout the start-up process, the motor maintains its position near the maximum torque-to-current ratio operating point, significantly improving current utilization and efficiency.

[0093] exist Figure 2 Based on open loop If The experimental waveform when the control is switched to dual closed-loop control is as follows: Figure 3 As shown, a load of 0.571 Nm is maintained before and after the switch. The control mode is switched when the switching flag changes from 0 to 1. It can be seen that the speed is stable without drop during the switch, the phase current is not distorted, and the peak current amplitude is 29.05 A. After the dual closed-loop steady state is switched, the phase current amplitude is 23.7 A. There is no obvious change in current amplitude before and after the switch, and the transition is smooth. Therefore, the adaptive start strategy of this invention can basically achieve a smooth switch between the two control methods.

[0094] The above description of the embodiments is provided to enable those skilled in the art to understand and apply the present invention. Those skilled in the art can readily make various modifications to the above embodiments and apply the general principles described herein to other embodiments without creative effort. Therefore, the present invention is not limited to the above embodiments, and any improvements and modifications made to the present invention by those skilled in the art based on the disclosure thereof should be within the scope of protection of the present invention.

Claims

1. A sensorless adaptive starting method for a high-speed permanent magnet synchronous motor, characterized in that, Includes the following steps: (1) According to If A mathematical model of a high-speed permanent magnet synchronous motor is developed for starting. A speed differential damper is designed to correct the reference speed and increase the system damping ratio. The compensation coefficient of the damper is then calculated and determined. k ; The expression for the rotational speed differential damper is as follows: in: ω i The rotating electric angular velocity of the stator current vector. ω i0 The reference speed before correction. Motor speed ω r The first derivative; The compensation coefficient k The calculation expression is as follows: in: K T The torque coefficient of the motor and K T =1.5 P n ψ f , J Let be the moment of inertia of the motor. P n This represents the number of pole pairs of the motor. I s0 and θ err0 These are the motors after they enter steady state. I s and θ err , I s The magnitude of the stator current vector. θ err for If The angle difference between the reference virtual coordinate system and the rotor position orientation coordinate system ψ f For permanent magnet flux linkage in motors; (2) Construct an extended state observer-phase-locked loop structure. Use this structure to demodulate rotor position and speed information while obtaining a smooth speed differential signal, providing feedback compensation basis for the speed differential damper. The expression for the extended state observer-phase-locked loop structure is as follows: in: Rotor position angle θ r The observed values, Motor speed ω r The observed values, For the observations of the extended state variables, , , They are respectively , , The first-order differential, β 1. β 2. β 3 represents the coefficients of the extended state observer and , e 1 represents the angle difference between the observed and actual rotor position angles. I s1 Given the initial value of the reference current, ω c To expand the bandwidth of the state observer; (3) Real-time detection feedback compensation If By referencing the angle difference between the virtual coordinate system and the rotor position orientation coordinate system, an adaptive starter controller is constructed, and the current vector magnitude and acceleration are adaptively adjusted based on MTPA. The q-axis reference current is also calculated and determined. i q_ref and reference rotor position angle θ ref Used to control the motor inverter.

2. The sensorless adaptive starting method for a high-speed permanent magnet synchronous motor according to claim 1, characterized in that, The angle difference e The expression for calculating 1 is as follows: in: e α and e β These are the α-axis back EMF and β-axis back EMF estimated by the sliding mode observer, respectively.

3. The sensorless adaptive starting method for a high-speed permanent magnet synchronous motor according to claim 1, characterized in that: The feedback compensation in step (3) If The angle difference between the reference virtual coordinate system and the rotor position orientation coordinate system ,in θ i In order to pass through ω i The electric angle obtained by integration, ω i The rotating electric angular velocity of the stator current vector is calculated using a rotational differential damper, i.e. .

4. The sensorless adaptive starting method for a high-speed permanent magnet synchronous motor according to claim 3, characterized in that, The reference speed ω i0 The calculation expression is as follows: in: α i This is the reference value for rotor angular acceleration. K p1 and K i1 Given the proportional coefficient and integral coefficient, b These are the weighting coefficients. t Indicates time.

5. The sensorless adaptive starting method for a high-speed permanent magnet synchronous motor according to claim 4, characterized in that, The weighting coefficient b The expression is as follows: in: ω ref Given the target rotational speed.

6. The sensorless adaptive starting method for a high-speed permanent magnet synchronous motor according to claim 3, characterized in that, In step (3), the q-axis reference current is determined. i q_ref and reference rotor position angle θ ref Specifically: when the observed motor speed Reach the target speed ω ref And when the motor enters a steady state, , Before this, , ; I s The magnitude of the stator current vector and , The closed-loop reference current amplitude is determined by... The difference is obtained after PI control. I comp This is the magnitude compensation amount for the stator current vector, i.e.: in: K p2 and K i2 Given the proportional coefficient and integral coefficient, b These are the weighting coefficients. t Indicates time.

7. The sensorless adaptive starting method for a high-speed permanent magnet synchronous motor according to claim 1, characterized in that, The specific method for controlling the motor inverter in step (3) is as follows: First, the three-phase stator current of the motor is collected and converted into the α-axis stator current. i α and β-axis stator current i β ,make θ ref As the coordinate transformation angle i α and i β Converted to d-axis stator current i d and q-axis stator current i q Then set the d-axis reference current. i d_ref =0, respectively for and PI control is used to obtain the d-axis stator voltage. u d and q-axis stator voltage u q ; Utilizing coordinate transformation angles θ ref Will u d and u q Converted to α-axis stator voltage u α and β-axis stator voltage u β Finally based on u α and u β A series of PWM signals are generated by SVPWM modulation to drive the power switching devices in the motor inverter.

Images