Double three-phase PMSM position sensorless control method based on random voltage injection

Abstract

The application discloses a double three-phase PMSM position sensorless control method based on random voltage injection, comprising the following steps: designing a novel random injection signal mixed with square waves and pulse waves, selecting a suitable signal period and adjusting the length ratio of the zero voltage section of the pulse wave signal; controlling the signal injection type and the injection target winding by adjusting the random number threshold parameter size to realize discontinuous differentiated injection; constructing an improved heterodyne method and a linear extended state observer to realize the rotor position and speed estimation method, feeding back the estimated rotor position and speed information to the vector control to realize the double three-phase permanent magnet synchronous motor position sensorless control. The application can effectively weaken the sharp noise caused by high-frequency voltage signal injection, significantly improve the estimation accuracy of the rotor position and speed of the double three-phase permanent magnet synchronous motor position sensorless control and the dynamic response characteristics of the system, and the injection signal algorithm structure is simple, the calculation burden is small, and it is easy to deploy on a low algorithm platform.

Description

Technical Field

[0001] This invention belongs to the field of motor control technology, specifically relating to a sensorless control method for dual three-phase PMSMs based on random voltage injection. Background Technology

[0002] Dual three-phase PMSMs (permanent magnet synchronous motors) have attracted widespread attention in high-reliability fields such as aerospace and surgical robots due to their advantages of high efficiency, low torque ripple, and high fault tolerance. However, in these applications, the aerospace field (such as aircraft electric propulsion systems) faces extreme vibrations and high temperatures under harsh operating conditions, while the surgical robot field is limited by complex electromagnetic interference caused by high-density equipment, leading to decreased sensor reliability and requiring additional space and volume for installation, making it difficult to meet the miniaturization requirements of the system. Certain applications lack the necessary installation conditions. Therefore, sensorless control has become an important research direction; in the zero-low speed region, high-frequency signal injection is currently the mainstream technical solution for achieving sensorless control.

[0003] In existing technologies, single-frequency high-frequency signal injection methods, such as those described in the literature ["A Sensorless Control Method for Zero-Speed ​​Region of Permanent Magnet Synchronous Motor Based on Stator Winding Neutral Point Voltage." IEEE Transactions on Industrial Applications, September / October 2016, Vol. 52, No. 5, pp. 4020-4028"], easily excite narrowband harmonics at the injection frequency, leading to torque pulsation and generating sharp electromagnetic noise audible to the human ear, severely affecting the equipment's operating experience. To mitigate this noise, traditional methods mainly include reducing the amplitude of the injected signal, adjusting the frequency of the injected signal, and using random frequency signal injection. However, reducing the amplitude of the injected signal deteriorates the signal-to-noise ratio and affects estimation accuracy; adjusting the frequency may lead to increased switching losses or signal overlap problems; and traditional random injection methods, such as the dual-frequency square wave combination method proposed in the literature [Comparative Study of Pseudo-random High-Frequency Signal Injection Schemes in Sensorless Built-in Permanent Magnet Synchronous Motor Drive System. IEEE Transactions on Power Electronics, March 2017, Vol. 32, No. 3, pp. 2123-2132], although broadening the spectrum, still retain discrete harmonic spikes in the power spectral density of its response current, resulting in limited noise suppression. In addition, the signal processing stage of existing methods has obvious shortcomings: the demodulation of traditional heterodyne methods is easily affected by voltage error interference introduced by factors such as inverter nonlinearity, leading to a decrease in position estimation accuracy; the widely used phase-locked loops have sluggish response under dynamic conditions such as load changes, resulting in large dynamic estimation errors. At the same time, some schemes use complex injected signals or observers to improve performance, which increases the difficulty of algorithm implementation and computational burden, making them unsuitable for deployment on low-computing-power control platforms in the above-mentioned applications.

[0004] In summary, existing sensorless control methods for dual three-phase permanent magnet synchronous motors still have significant shortcomings in terms of sharp noise suppression, rotor position estimation accuracy, system dynamic robustness, and ease of engineering implementation. There is an urgent need to design a sensorless control scheme for dual three-phase permanent magnet synchronous motors that balances low noise, high accuracy, and strong robustness. Summary of the Invention

[0005] In view of the above, the present invention provides a position sensorless control method for dual three-phase PMSM based on random voltage injection. This method achieves high-precision estimation of rotor position and speed by means of an innovative hybrid signal injection strategy and a highly robust observer architecture design, while significantly suppressing sharp electromagnetic noise, and greatly improving the dynamic response performance and operating condition adaptability of the system.

[0006] A sensorless control method for dual three-phase PMSMs based on random voltage injection includes the following steps:

[0007] (1) Design a coordinated injection strategy for selecting the injection signal type and winding of a dual three-phase PMSM to generate high-frequency random injection voltage signals for two sets of windings. u inj1 and u inj2 ;

[0008] (2) Collect the three-phase stator currents of the two sets of windings and transform them to their respective αβ coordinate systems, and then calculate the combined α-axis current. i α and β-axis current i β ;

[0009] (3) Design an improved heterodyne method and a linear extended state observer (LESO) algorithm to estimate the rotor position and speed, thereby estimating the rotor position angle. and rotational speed ;

[0010] (4) Inject high-frequency random voltage signals u inj1 and u inj2 and rotor position angle and rotational speed It is incorporated into the vector control of the motor to achieve sensorless closed-loop control of dual three-phase PMSM.

[0011] Furthermore, the specific implementation of step (1) is as follows: First, define the following two high-frequency injection voltage signals:

[0012]

[0013] in:u 01 ( t )express t High-frequency pulse signal at time, u 02 ( t )express t The high-frequency square wave signal at time t, T p1 and T p2 These are the periods of the high-frequency pulse signal and the high-frequency square wave signal, respectively. U inj The voltage amplitude of the injected signal. t r1 ( t )and t r2 ( t () indicates time t Divide by respectively T p1 and T p2 The remainder;

[0014] Then determine the type of injection signal: generate a random number between 0 and 1 in each injection signal cycle. Q and S ,like Q Less than the set value Q 0 indicates the high-frequency injection voltage signal is determined. u inj_all ( t () is a high-frequency pulse signal u 01 ( t );like Q Greater than or equal to Q 0 indicates the high-frequency injection voltage signal is determined. u inj_all ( t () is a high-frequency square wave signal u 02 ( t );

[0015] Finally, determine the injection signal for the winding: If S Less than the set value S 0, then a high-frequency random voltage signal is injected into winding 1. u inj1 for u inj_all ( t High-frequency random injection voltage signal for winding No. 2 u inj2 If it is 0; S Greater than or equal to the set value S0, then a high-frequency random voltage signal is injected into winding 1. u inj1 High-frequency random injection voltage signal for windings 0 and 2. u inj2 for u inj_all ( t ).

[0016] Furthermore, in step (2), the synthesized α-axis current is calculated using the following formula. i α and β-axis current i β :

[0017]

[0018]

[0019] in: and These are the α-axis stator current and β-axis stator current of winding 1, respectively. and These are the α-axis stator current and β-axis stator current of winding 2, respectively. T This is the coordinate transformation matrix from the αβ coordinate system of winding 2 to the αβ coordinate system of winding 1.

[0020] Furthermore, in step (3), the rotor position angle is estimated using the following observer expression. and rotational speed ;

[0021]

[0022] in: , , The state variables of the observer and , , , This represents the load torque of the motor. , , Corresponding to , , The first derivative, This represents the number of pole pairs of the motor. J Let be the moment of inertia of the motor. This is an estimated value for the electromagnetic torque of the motor. , , The gain coefficient of the observer. e This represents the observation error regarding the rotor position angle.

[0023] Furthermore, the observation error e The expression is as follows:

[0024]

[0025]

[0026] ,

[0027] ,

[0028]

[0029] in: and These are the sine and cosine components of the orthogonal signal related to the rotor position, respectively. , , , This represents the difference in current components at adjacent signal demodulation moments. i α(0) , i α(1) , i α(2) They represent t r 0, T p2 / 2、 T p2 α-axis current at time i β(0) , i β(1) , i β(2) They represent t r 0, T p2 / 2、 T p2 β-axis current at time K 2 is an intermediate variable. t r Indicates time t Divide by T p The remainder, T p Injecting voltage signals at high frequencies u inj_all ( t The cycle of ) and These are the d-axis inductance and q-axis inductance of the motor, respectively.

[0030] Furthermore, the electromagnetic torque estimate The expression is as follows:

[0031]

[0032]

[0033] in: and These are the coordinates corresponding to the synthesized αβ coordinate system. shaft current estimates and shaft current estimate and These are the d-axis inductance and q-axis inductance of the motor, respectively. This refers to the permanent magnet flux linkage of the motor.

[0034] Furthermore, the gain coefficient , , The expression is as follows:

[0035]

[0036] in: ω 0 represents the observer bandwidth.

[0037] Furthermore, the specific implementation of step (4) is as follows:

[0038] First, utilize the angle For the α-axis stator current of winding 1 and β-axis stator current Perform coordinate transformation to obtain winding number 1. shaft current estimate and shaft current estimate ; Utilizing angles For the α-axis stator current of winding No. 2 and β-axis stator current Perform coordinate transformation to obtain winding number 2. shaft current estimate and shaft current estimate ;

[0039] Then to PI (Proportional-Integral) control is used to obtain the q-axis current reference value. ,Will The result obtained through PI control plus u inj1 get Shaft voltage reference value ,right PI control is performed to obtain Shaft voltage reference value ,Will The result obtained through PI control plus u inj2 get Shaft voltage reference value ,right PI control is performed to obtain Shaft voltage reference value ;in and These are the No. 1 windings. Shaft current reference value and Shaft current reference value, and These are the No. 2 windings. Shaft current reference value and Shaft current reference value, , , This is a reference value for motor speed;

[0040] Finally, use angles right and The coordinate transformation is performed to obtain the α-axis voltage reference value. and β-axis voltage reference value Then, PWM signals are generated using SVP technology to drive the switching devices in the inverter corresponding to winding 1; utilizing angle right and The coordinate transformation is performed to obtain the α-axis voltage reference value. and β-axis voltage reference value Then, PWM signals are generated using SVP technology to drive the switching devices in the inverter corresponding to the No. 2 winding.

[0041] 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 control method for dual three-phase PMSMs based on random voltage injection.

[0042] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described sensorless control method for dual three-phase PMSMs based on random voltage injection.

[0043] Based on the above technical solution, the present invention has the following beneficial technical effects:

[0044] 1. The present invention designs a signal type and winding selection-based coordinated injection strategy based on random injection of square wave and pulse wave. Taking into account the characteristic that the two sets of windings of the dual three-phase permanent magnet synchronous motor can be controlled independently, the invention significantly suppresses harmonic spikes in PSD (power spectral density) by optimizing the injection waveform combination, reasonably selecting the injection frequency and related random number threshold parameters, and significantly reducing the measurement results of sharp noise level, making it more suitable for application scenarios such as surgical robots.

[0045] 2. This invention proposes a rotor position and speed estimation algorithm that integrates an improved heterodyne method and a linear extended state observer. Under various operating conditions, the algorithm has higher estimation accuracy, stronger robustness and dynamic response capability. Compared with existing schemes, the maximum speed error and peak-to-peak position error are reduced by more than 23%, verifying its effectiveness in positionless control of dual three-phase permanent magnet synchronous motors. Attached Figure Description

[0046] Figure 1 This is a block diagram of a dual three-phase PMSM sensorless control system based on random voltage injection, according to an embodiment of the present invention.

[0047] Figure 2 This is a schematic diagram of the signal generator module in an embodiment of the present invention.

[0048] Figure 3 This is a flowchart illustrating the coordinated injection strategy of signal type and winding selection in an embodiment of the present invention.

[0049] Figure 4 This is a flowchart illustrating the signal processing (demodulation and observation) part in an embodiment of the present invention.

[0050] Figure 5 This diagram illustrates the A1 phase current and its amplified waveform, as well as the PSD analysis results of the A1 phase current, when a dual three-phase permanent magnet synchronous motor operates at 200 r / min and 100% rated load, using the method of this invention and the traditional signal injection method. In the diagram, (a) to (f) correspond to the 2.5 kHz square wave voltage injection method, the 1.67 kHz square wave voltage injection method, the 1.67 kHz pulse wave voltage injection method, the 2.5 kHz square wave and 1.25 kHz square wave voltage mixed injection method, the 2.5 kHz square wave and 1.67 kHz square wave voltage mixed injection method, and the 2.5 kHz square wave and 1.67 kHz pulse wave voltage mixed injection method.

[0051] Figure 6 The diagram shows a simulation comparison of phase current, speed, speed error, and angle error under the method of this invention and the traditional signal processing method when the dual three-phase permanent magnet synchronous motor is running at 100% rated load and with a speed variation of 200-0-200 r / min. In the diagram, (a) and (b) correspond to the traditional signal processing method and the method of this invention, respectively.

[0052] Figure 7 The diagram shows a simulation comparison of the speed, speed error, and angle error of a dual three-phase permanent magnet synchronous motor operating at 200 r / min with a load change of 0-100%-0 rated load, using the method of this invention and the traditional signal processing method. In the diagram, (a) and (b) correspond to the traditional signal processing method and the method of this invention, respectively. Detailed Implementation

[0053] 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.

[0054] like Figure 1 As shown, this embodiment provides a position-sensorless control method for dual three-phase PMSMs based on random voltage injection, including the following steps:

[0055] (1) Design a coordinated injection strategy for selecting the injection signal type and winding of a dual three-phase permanent magnet synchronous motor. First, a novel hybrid random injection voltage signal is constructed, which is a random combination of an integer-cycle pulse signal and a square wave signal; then, the random number threshold parameter is adjusted. S 0 and Q The magnitude of 0 controls the proportion of the pulse signal in the mixed random injection voltage signal and the proportion of the mixed random injection voltage signal injected into the A1B1C1 winding, thereby selectively injecting the generated mixed high-frequency voltage signal into the estimated coordinate axis of the vector control of the two independent windings (A1B1C1 winding and A2B2C2 winding) of the dual three-phase permanent magnet synchronous motor.

[0056] A novel random injection signal is constructed by randomly combining pulse and square wave signals. The pulse signal consists of a square wave signal superimposed with a zero-voltage signal. The duration of the zero-voltage signal serves as an adjustable degree of freedom, allowing for flexible adjustment of the pulse signal frequency. To ensure signal-to-noise ratio, the square wave portion of the full-cycle pulse signal maintains the same injection period and amplitude as the used full-cycle square wave signal. To ensure the quality of the injected signal and achieve system synchronization, the periods of both designed high-frequency injection signals are set to integer multiples of the PWM carrier signal period. u 01 , u 02 These are a pulse signal and a square wave signal, each with a full cycle, and their periods are as follows:

[0057]

[0058] in: T p1 , Tp2 , T pwm These are the periods of the pulse signal, square wave signal, and PWM carrier signal, respectively. f p1 , f p2 The frequency of the pulse signal and square wave signal (is a set value and) f p1 < f p2 ); a 1. a 2 represents the ratio of the period of the pulse signal to the period of the carrier signal, and the period of the square wave signal to the period of the carrier signal.

[0059] The expression for two injected signals is written as:

[0060]

[0061] in: U inj The amplitude of the injected signal (set value); t r1 , t r2 They represent time respectively t Divide by T p1 , T p2 The remainder. This implementation will design a novel signal generator based on the two basic signals mentioned above, such as... Figure 2 As shown, this is to further enhance the discontinuity of the injected signal.

[0062] To enhance the randomness and discontinuity of the injected signals from the two windings, thereby effectively suppressing the discrete spectral components in the power spectral density of the response current and reducing the amplitude of the high-frequency response current, the novel signal generator designed in this embodiment mainly consists of a signal selection section and a winding selection section, such as... Figure 3 As shown, the method consists of the following four steps:

[0063] ① Set two threshold parameters between 0 and 1, denoted as Q 0 and S 0 is used to control the signal type and the selection probability of the injection winding, respectively; two random numbers between 0 and 1 are generated using the linear congruential method, denoted as . Q and S .

[0064] ②Based on random numbers Q and threshold Q The comparison result of 0 selects the injection signal type: if Q < Q0, inject a pulse signal; otherwise, inject a square wave signal, both lasting for a full cycle.

[0065] ③Based on random numbers S and threshold S The comparison result of 0 selects the signal injection winding: if S < S 0, the signal is injected into winding A1B1C1; otherwise, it is injected into winding A2B2C2.

[0066] ④ After the full cycle signal injection is completed, the system will regenerate random numbers and execute the subsequent injection process in a loop according to the preset procedure.

[0067] Ultimately, two sets of winding injection signals are generated, among which... u inj_all For voltage signals that only participate in the signal type selection section, u inj1 , u inj2 These are the injected signals on windings A1B1C1 and A2B2C2, respectively; the expression for these signals can be written as:

[0068]

[0069] in: R all , R 1. R 2 are random operators for the signal that only participates in the signal type selection part, the signal injected into the A1B1C1 winding, and the signal injected into the A2B2C2 winding, respectively. r This indicates that the current number is the [number]. r Injection cycle of one whole cycle.

[0070] (2) Collect the high-frequency response current of the two sets of windings and convert them to a unified composite coordinate system to obtain the composite high-frequency response current.

[0071] When the motor is running in the zero-speed range, since the frequency of the injected signal is much higher than the fundamental frequency of the motor, the effects of resistance, back electromotive force, and cross-coupling terms in the motor can be ignored. The high-frequency mathematical model of the motor can be expressed as:

[0072]

[0073] in: s It is a differential operator; u dh , u qh The high-frequency component of the stator voltage in the dq coordinate system (a rotating coordinate system with the rotor magnetic field position orientation); i dh , iqh This corresponds to the high-frequency current component; L dh , L qh This corresponds to the high-frequency incremental inductance; the subscript h indicates the high-frequency component. There is also... θ e1 and θ e2 These are the rotor position electrical angles based on the α1 and α2 axis position references, respectively. θ e1 = θ e2 +π / 6; and , and These represent the estimated rotor position electrical angle value obtained after injecting high-frequency signals into windings A1B1C1 and A2B2C2, the estimated rotor position electrical angle value, and the electrical angle error between the estimated rotor position electrical angle and the actual rotor position electrical angle, respectively. , , This represents the actual rotor position angle.

[0074] Define the rotation transformation matrix as:

[0075]

[0076] in: θ The angle used for rotational transformation. For winding A1B1C1, in - When a random signal is injected into the coordinate system (a rotating coordinate system established based on the winding under the position orientation of the rotor magnetic field estimation A1B1C1), the high-frequency current response components in the dq coordinate system are written as:

[0077]

[0078] correspond - The high-frequency response current components in the coordinate system are:

[0079]

[0080] According to the above formula, when a signal is injected into the A1B1C1 winding... u inj1 At that time, the high-frequency response current in the α1-β1 coordinate system (a stationary coordinate system established based on the A1B1C1 winding) is:

[0081]

[0082] The current in the A2B2C2 winding, located in the α2-β2 coordinate system (a stationary coordinate system established based on the A2B2C2 winding), is transformed to the α1-β1 coordinate system corresponding to the A1B1C1 winding using coordinate transformation to construct a synthetic high-frequency current component. This component is then used to extract rotor position and speed information through a unified signal processing flow. The direction of the synthetic coordinate system coincides with the direction of the A1B1C1 winding coordinate system. - The coordinate system is obtained from the α2-β2 coordinate system through coordinate transformation, and the two differ by 30°. - The high-frequency current component in the coordinate system can be expressed as:

[0083]

[0084] When a signal is injected into the A2B2C2 winding u inj2 hour, - The high-frequency response current in the coordinate system is:

[0085]

[0086] Since the control structures of the two windings are identical, it can be assumed that their estimation errors are the same. At this point, the high-frequency response current in the synthesized α-β coordinate system (a stationary coordinate system in the same direction as the α1-β1 coordinate system) can be expressed as:

[0087]

[0088] (3) Constructing a rotor position and speed estimation method that integrates the improved heterodyne method with the linear extended state observer: First, the synthesized high-frequency response current is demodulated using the improved heterodyne method to extract the equivalent error signal containing rotor position information. Then, the rotor position and speed of the motor are estimated in real time using the linear extended state observer, and the estimated rotor position and speed information is fed back to the motor vector control to achieve a low-noise, high-precision, and robust sensorless closed-loop control of the dual three-phase permanent magnet synchronous motor.

[0089] When considering the effects of inverter nonlinear characteristics, stator resistance voltage drop, and other factors, within a complete square wave signal injection cycle... - The actual injected high-frequency voltage in the coordinate system is:

[0090]

[0091] in: and They are respectively shaft and The error voltage on the shaft, Δ, is the error voltage caused by changes in the vector control (FOC) command due to sudden changes in operating conditions. u FOC Error voltage caused by inverter dead time effect u DT Together constitute; Δ u FOC The sudden change in motor operating conditions (such as a load jump or rapid change in speed) causes a significant adjustment in the vector control output command, which is equivalent to the fluctuation of the injected high-frequency voltage signal. u DT This is caused by nonlinear voltage errors due to the inverter dead-zone effect when the direction or amplitude of the motor phase current changes, resulting in distortion of the high-frequency injected signal.

[0092] To minimize the aforementioned errors, the method sets the current sampling period, signal demodulation period, and FOC control period to be consistent, and both to be half the period of the square wave signal. In this case, the FOC command remains constant within the signal injection period, thus ensuring Δ... u FOC The dead-zone voltage error remains unchanged; meanwhile, since the amplitudes of the positive and negative half-cycle voltage vectors injected by the square wave are equal and their directions are opposite, the dead-zone voltage error between the demodulation times of the two continuous signals remains unchanged. u DT It can be considered to remain unchanged. Under the above conditions, the synthesized error voltage within a unit injection cycle is a constant value. Therefore, the error value Δ obtained by demodulating the high-frequency response current in the synthesized α-β coordinate system using the general heterodyne method is... i α Δ i β It can be represented as:

[0093]

[0094] Where: Δ T This is the signal demodulation period.

[0095] Because Δ i α Δ i β The current difference is severely affected by error voltage, leading to a decrease in rotor position estimation accuracy. To suppress the deterioration of rotor position estimation accuracy caused by output voltage error, an improved heterodyne method based on the difference between two current changes is designed, such as... Figure 4 As shown, after normalizing the synthetic high-frequency response current containing position information, the resulting position error signal is:

[0096]

[0097]

[0098] In the formula: i α(0) , i α(1) , i α(2) They are respectively t r1 or t r2 0, T p2 / 2、 T p2 The α-axis current value at that time i β(0) , i β(1) , i β(2) They are respectively t r1 or t r2 0, T p2 / 2、 T p2 The β-axis current value at that time; Δ i cos Δ i sin These represent the cosine and sine signals related to the rotor position, respectively. η As a variable, it represents:

[0099]

[0100] The above signal demodulation scheme can effectively eliminate voltage error components in orthogonal response signals that contain location information.

[0101] After obtaining the equivalent position error, a linear extended state observer was designed to estimate the rotor position and speed in real time to improve the overall system performance. This linear state observer uses the electromagnetic torque estimate as a feedforward input, and based on this, the control objective is achieved by constructing a motor position estimation loop. The corresponding extended state equation can be written as:

[0102]

[0103] in: , , , , For load torque, This represents the number of pole pairs of the motor. J For rotational inertia, , , The gain coefficient of the observer. The estimated value of electromagnetic torque can be written as:

[0104]

[0105]

[0106] in: and These are the coordinates corresponding to the synthesized αβ coordinate system. shaft current estimates and shaft current estimate and These are the d-axis inductance and q-axis inductance of the motor, respectively. This refers to the permanent magnet flux linkage of the motor.

[0107] To ensure system stability, the three poles in the characteristic equation can be located at the same location. ω The points at point 0 (i.e., the observer's bandwidth) coincide, and are located at... s In the left half-plane of the plane, the gain coefficient is selected as follows:

[0108]

[0109] When the position error converges to zero, it means that the estimated position angle is consistent with the actual position angle of the rotor. At this time, the estimated position angle and speed output by the linear expansion state observer provide feedback for vector control, thereby realizing positionless control of the dual three-phase permanent magnet synchronous motor.

[0110] To verify the effectiveness and superiority of the sensorless control method for dual three-phase permanent magnet synchronous motors based on random voltage injection, we conducted simulation verification. During the simulation, the control cycle of the algorithm was 50µs. Taking a dual three-phase built-in permanent magnet synchronous motor as an example, the parameters are shown in Table 1:

[0111] Table 1

[0112]

[0113] Based on the motor parameters in the table above, the voltage frequency and amplitude parameters for various injection methods are designed as shown in Table 2:

[0114] Table 2

[0115]

[0116] To verify the high-frequency noise suppression performance of the positionless control method of the present invention, when the motor is running at a speed of 200 r / min and 100% rated load (12 N·m), the random value threshold parameter is... Q 0 and SAll values ​​were set to 0.5, and different types of voltage signals were injected to achieve a comparative simulation of motor positionless control. For example... Figure 5 The simulation results of various injection methods at 200 r / min and 12 N·m are shown. The first and second rows of the figure show the A1 phase current waveforms, and the third row shows the power spectral density analysis results of the A1 phase current under the corresponding injection method. As shown in the A1 phase current waveform in the second row of each subfigure, when different voltage signals are injected, the A1 phase current will generate a response signal with the same frequency as the injected signal. As shown in the power spectral density analysis results in the third row of each subfigure, under the conditions of single square wave, single pulse wave, and combination of two square waves, obvious discrete harmonic peaks appear in the power spectral density at the least common multiple and odd multiples of the injection frequency. However, under the collaborative injection scheme of signal type and winding selection based on square wave and pulse wave of this invention, no such discrete harmonics appear in the power spectral density results, indicating that the scheme can effectively suppress discrete harmonics in the power spectral density of the response current, thereby significantly reducing high-frequency sharp noise.

[0117] To verify the dynamic performance of the positionless control method of this invention, simulation results were obtained by measuring the A1 and A2 phase currents, actual speed and estimated speed, the error between actual and estimated speed, and the error between actual and estimated angle under the improved signal demodulation method of this invention and the traditional signal demodulation method, respectively, when the motor was running at 100% rated load (12 N·m) with speed changes of 200-0-200 r / min and 200 r / min with load changes of 0-100%-0 rated load. Figure 6 The simulation results are shown for a speed of 12 N·m with a rotational speed variation of 200-0-200 r / min. The first row of the figure shows the current waveforms of phases A1 and A2; the second row shows the actual and estimated rotational speed waveforms; the third row shows the error between the actual and estimated rotational speeds; and the fourth row shows the error between the actual and estimated angles. As can be seen from the simulation results in the third and fourth rows of each sub-figure, the improved signal demodulation scheme of this invention exhibits advantages in dynamic performance under both driving and braking modes. Specifically, the maximum rotational speed error is reduced from 35.83 r / min to 23.71 r / min, a reduction of 33.8%; the peak-to-peak position error is reduced from 8.69° to 5.30°, a reduction of 39.0%. Figure 7The simulation results are shown for a load variation of 0-100%-0 rated load at 200 r / min. The first row shows the waveforms of the actual and estimated rotational speeds, the second row shows the errors of the actual and estimated rotational speeds, and the third row shows the errors of the actual and estimated angles. As can be seen from the simulation results in the second and third rows of each sub-figure, the improved signal demodulation scheme of this invention has stronger robustness and better dynamic performance during sudden load increases and decreases. The maximum rotational speed errors of the traditional signal demodulation scheme under this condition are 32.96 r / min and 30.05 r / min, respectively, with corresponding peak-to-peak dynamic position errors of 8.17° and 10.04°. In contrast, the rotational speed errors of the method of this invention are reduced to 12.25 r / min and 14.21 r / min, respectively, representing reductions of 62.9% and 52.7%; the peak-to-peak position errors are reduced to 6.26° and 6.12°, respectively, representing reductions of 23.4% and 39.0%.

[0118] 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 control method for dual three-phase PMSMs based on random voltage injection, characterized in that, Includes the following steps: (1) Design a coordinated injection strategy for selecting the injection signal type and winding of a dual three-phase PMSM to generate high-frequency random injection voltage signals for two sets of windings. u inj1 and u inj2 ; (2) Collect the three-phase stator currents of the two sets of windings and transform them to their respective αβ coordinate systems, and then calculate the combined α-axis current. i α and β-axis current i β ; (3) Design an improved rotor position and speed estimation algorithm that integrates the heterodyne method with the linear extended state observer, so as to estimate the rotor position angle through the following observer expression. and rotational speed ; in: , , The state variables of the observer and , , , This represents the load torque of the motor. , , Corresponding to , , The first derivative, This represents the number of pole pairs of the motor. J Let be the moment of inertia of the motor. This is an estimated value for the electromagnetic torque of the motor. , , The gain coefficient of the observer. e This refers to the observation error regarding the rotor position angle; (4) Inject high-frequency random voltage signals u inj1 and u inj2 and rotor position angle and rotational speed It is incorporated into the vector control of the motor to achieve sensorless closed-loop control of dual three-phase PMSM.

2. The position-sensorless control method for dual three-phase PMSM based on random voltage injection according to claim 1, characterized in that, The specific implementation method of step (1) is as follows: First, define the following two high-frequency injection voltage signals: in: u 01 ( t )express t High-frequency pulse signal at time, u 02 ( t )express t The high-frequency square wave signal at time t, T p1 and T p2 These are the periods of the high-frequency pulse signal and the high-frequency square wave signal, respectively. U inj The voltage amplitude of the injected signal. t r1 ( t )and t r2 ( t () indicates time t Divide by respectively T p1 and T p2 The remainder; Then determine the type of injection signal: generate a random number between 0 and 1 in each injection signal cycle. Q and S ,like Q Less than the set value Q 0 indicates the high-frequency injection voltage signal is determined. u inj_all ( t () is a high-frequency pulse signal u 01 ( t );like Q Greater than or equal to Q 0 indicates the high-frequency injection voltage signal is determined. u inj_all ( t () is a high-frequency square wave signal u 02 ( t ); Finally, determine the injection signal for the winding: If S Less than the set value S 0, then a high-frequency random voltage signal is injected into winding 1. u inj1 for u inj_all ( t High-frequency random injection voltage signal for winding No. 2 u inj2 If it is 0; S Greater than or equal to the set value S 0, then a high-frequency random voltage signal is injected into winding 1. u inj1 High-frequency random injection voltage signal for windings 0 and 2. u inj2 for u inj_all ( t ).

3. The position-sensorless control method for dual three-phase PMSM based on random voltage injection according to claim 1, characterized in that, In step (2), the synthesized α-axis current is calculated using the following formula. i α and β-axis current i β : in: and These are the α-axis stator current and β-axis stator current of winding 1, respectively. and These are the α-axis stator current and β-axis stator current of winding 2, respectively. T This is the coordinate transformation matrix from the αβ coordinate system of winding 2 to the αβ coordinate system of winding 1.

4. The position-sensorless control method for dual three-phase PMSM based on random voltage injection according to claim 2, characterized in that, The observation error e The expression is as follows: ， ， in: and These are the sine and cosine components of the orthogonal signal related to the rotor position, respectively. , , , This represents the difference in current components at adjacent signal demodulation moments. i α(0) , i α(1) , i α(2) They represent t r 0, T p2 / 2、 T p2 α-axis current at time i β(0) , i β(1) , i β(2) They represent t r 0, T p2 / 2、 T p2 β-axis current at time K 2 is an intermediate variable. t r Indicates time t Divide by T p The remainder, T p Injecting voltage signals at high frequencies u inj_all ( t The cycle of ) and These are the d-axis inductance and q-axis inductance of the motor, respectively.

5. The position-sensorless control method for dual three-phase PMSM based on random voltage injection according to claim 1, characterized in that, The electromagnetic torque estimate The expression is as follows: in: and These are the coordinates corresponding to the synthesized αβ coordinate system. shaft current estimates and shaft current estimate and These are the d-axis inductance and q-axis inductance of the motor, respectively. This refers to the permanent magnet flux linkage of the motor.

6. The position-sensorless control method for dual three-phase PMSM based on random voltage injection according to claim 1, characterized in that, The gain coefficient , , The expression is as follows: in: ω 0 represents the observer bandwidth.

7. The position-sensorless control method for dual three-phase PMSM based on random voltage injection according to claim 1, characterized in that, The specific implementation method of step (4) is as follows: First, utilize the angle For the α-axis stator current of winding 1 and β-axis stator current Perform coordinate transformation to obtain winding number 1. shaft current estimate and shaft current estimate ; Utilizing angles For the α-axis stator current of winding No. 2 and β-axis stator current Perform coordinate transformation to obtain winding number 2. shaft current estimate and shaft current estimate ; Then to PI control is used to obtain the q-axis current reference value. ,Will The result obtained through PI control plus u inj1 get Shaft voltage reference value ,right PI control is performed to obtain Shaft voltage reference value ,Will The result obtained through PI control plus u inj2 get Shaft voltage reference value ,right PI control is performed to obtain Shaft voltage reference value ;in and These are the No. 1 windings. Shaft current reference value and Shaft current reference value, and These are the No. 2 windings. Shaft current reference value and Shaft current reference value, , , This is a reference value for motor speed; Finally, use angles right and The coordinate transformation is performed to obtain the α-axis voltage reference value. and β-axis voltage reference value Then, PWM signals are generated using SVP technology to drive the switching devices in the inverter corresponding to winding 1; utilizing angle right and The coordinate transformation is performed to obtain the α-axis voltage reference value. and β-axis voltage reference value Then, PWM signals are generated using SVP technology to drive the switching devices in the inverter corresponding to the No. 2 winding.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: The processor is used to execute the computer program to implement the position-sensorless control method for dual three-phase PMSM based on random voltage injection as described in any one of claims 1 to 7.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by the processor, it implements the position-sensorless control method for dual three-phase PMSM based on random voltage injection as described in any one of claims 1 to 7.

Images