A full speed range sensorless control method for synchronous reluctance machines

Abstract

This invention discloses a sensorless control method for the entire speed domain of a synchronous reluctance motor. Addressing the issue of nonlinear fluctuations in the angle signal affecting motor operational stability, it employs a smoothing approach for the error signal, resolving the abrupt speed jumps caused by varying estimation errors in the sensorless algorithm across the entire speed domain. A phase-locked loop (PLL) is used to fuse the error signals during the speed transition range, and the error weights are adjusted based on the system operating speed to modify the error fed back to the PLL. This invention uses the position error signal as the PLL tracking signal to construct a composite observer. This method eliminates oscillations in the motor speed estimation caused by purely linear changes in the weighting coefficients, thereby reducing the risk of motor step loss and improving operational stability.

Description

Technical Field

[0001] This invention relates to the field of power electronics technology, and in particular to a sensorless control method for the full speed range of synchronous reluctance motors. Background Technology

[0002] Synchronous reluctance motors (SRRMs) are widely used in many fields due to their advantages such as wide speed range, low production cost, low torque ripple, and simple and reliable structure. In high-performance vector control of SRRMs, precise rotor position information is required for closed-loop control of the current and speed loops, as well as coordinate transformation. However, introducing position sensors reduces system reliability, and special applications such as high temperature, vibration, and long distances limit the use of sensors. Therefore, achieving sensorless control of SRRMs is particularly urgent.

[0003] In the full speed range of a motor, a single sensorless control strategy is insufficient for accurate estimation. Therefore, the full speed range of the motor is typically divided into two operating stages: zero-speed and medium-speed, with different algorithms used at each speed. In the zero-speed stage, the rotor position is determined using the traditional high-frequency voltage injection method. Algebraic operations are used to separate the fundamental and high-frequency components in the current response, reducing observer phase lag and improving system dynamic performance. In the medium-speed stage, a sliding mode control law is designed using a saturated function with smooth and continuous characteristics. An adaptive observer is designed using the feedback loop of the back electromotive force, and a phase-locked loop is used to compensate for the rotor phase delay, improving the system robustness and observation accuracy.

[0004] The key to estimating the motor's position across its full speed range using different algorithms is ensuring a smooth transition between the two sensorless strategies during the switching interval. In traditional weighted composite control, the weighting coefficients increase or decrease linearly during the switching interval. Because the two sensorless algorithms have different estimation errors, a simple linear change in the weighting coefficients cannot eliminate the oscillations in the system's estimation error, leading to abrupt speed jumps. To address this problem, a novel composite observer is designed. The core idea is to abandon the use of angle signals and instead smooth the error signals. Within the transition interval, the position error signals from the two control strategies are fused, and the error weights are adjusted according to the operating speed to change the error amount in the feedback loop. Summary of the Invention

[0005] The purpose of this invention is to provide a full-speed-domain sensorless control method for synchronous reluctance motors, which solves the problems of motor speed estimation oscillation and motor step loss caused by the linear change of the weight coefficient of the traditional observer.

[0006] The technical solution to achieve the objective of this invention is as follows: Firstly, this invention provides a sensorless control method for the entire speed range of a synchronous reluctance motor, comprising the following steps:

[0007] Step 1: Construct a mathematical model of a synchronous reluctance motor under high-frequency excitation;

[0008] Step 2: Select the high-frequency signal injection strategy in the transition range. When the motor speed is in the zero low speed range, the high-frequency injection method is used to obtain rotor information. At this time, a high-frequency square wave signal with an amplitude of one-tenth of the DC bus voltage needs to be injected into the system. The range of the injected high-frequency signal should not exceed the upper speed limit switching point.

[0009] Step 3, design of composite position observer based on fused phase-locked loop; first, the phase-locked loop of high frequency injection method and sliding mode algorithm is merged, the position estimation information of the two is subtracted from the final position of composite observer and fed back to phase-locked loop, the position errors of the two control strategies are weighted and fused in the transition interval, and the error weight is adjusted according to the motor running speed to change the amount of error fed back to phase-locked loop;

[0010] Step 4: Calculate the transition interval switching point;

[0011] Step 5, selecting the weighting function of the composite observer; after merging the phase-locked loops of the high-frequency injection method and the sliding mode algorithm, the weighting coefficients of the error signal are determined using the speed switching point calculated in Step 4; when the speed is less than the lower limit switching point, the position error signal is entirely provided by the difference between the high-frequency injection method and the composite observer; if the speed is higher than the upper limit switching point, the error is entirely from the difference between the sliding mode algorithm and the composite observer; when the speed is in the transition range, the error is composed of both the high-frequency injection method and the sliding mode algorithm.

[0012] In a second aspect, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the method described in the first aspect.

[0013] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in the first aspect.

[0014] Fourthly, the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method described in the first aspect.

[0015] Compared with the prior art, the significant advantages of this invention are:

[0016] The transition algorithm of the present invention, which combines the sensorless control scheme of the motor in the low-speed and medium-speed ranges, can improve the tracking accuracy of the composite observer and the system's resistance to load disturbances, while reducing the phase delay and system chattering caused by the extensive use of filters. Attached Figure Description

[0017] Figure 1 This is a block diagram of a composite position observer structure based on a fused phase-locked loop in an embodiment of the present invention.

[0018] Figure 2 This is a flowchart of the composite observer in an embodiment of the present invention.

[0019] Figure 3 This is a block diagram of a sensorless composite control structure for a synchronous reluctance motor across its entire speed range.

[0020] Figure 4 This is a speed response curve of an improved composite position observer based on a fused phase-locked loop.

[0021] Figure 5 This is a comparison diagram of rotor positions of an improved composite position observer based on a fused phase-locked loop. Detailed Implementation

[0022] A composite position observer based on a fused phase-locked loop (PLL) for a synchronous reluctance motor weighted and fused the estimation errors of two sensorless control strategies within the transition range. The error weights were adjusted according to the operating speed to change the amount of error fed back to the PLL. Its block diagram is shown below. Figure 1 As shown.

[0023] This invention provides a sensorless control method for the entire speed domain of a synchronous reluctance motor, the derivation flowchart of which is shown below. Figure 2 As shown, it includes the following steps:

[0024] Step 1: Derivation of the mathematical model of the motor under high-frequency excitation. The mathematical model of SynRM under high-frequency excitation differs from the basic mathematical model. The high-frequency signal injection method relies on accurate model parameters and precise extraction of high-frequency signal feedback. Therefore, a mathematical model of the synchronous reluctance motor under high-frequency excitation needs to be built as the research basis. The high-frequency model follows the assumptions of the traditional ideal motor model.

[0025] Step 2: Selection of high-frequency signal injection strategy in the transition range. When the motor speed is in the zero low speed range, the high-frequency injection method is used to obtain rotor information. At this time, a high-frequency square wave signal with an amplitude of one-tenth of the DC bus voltage needs to be injected into the system. The range of the injected high-frequency signal should not exceed the upper limit switching point of the speed.

[0026] Step 3: Design of a composite position observer based on a fused phase-locked loop (PLL). First, the PLLs of the high-frequency injection method and the sliding mode algorithm are merged. The difference between the position estimates from both methods and the final position of the composite observer is fed back to the PLL. During the transition interval, the position errors of the two control strategies are weighted and fused, and the error weights are adjusted according to the motor speed to change the amount of error fed back to the PLL. The core idea of ​​the proposed algorithm is to abandon the use of angle signals and instead smooth the error signals.

[0027] Step 4: Calculation of the transition interval switching points. The switching points ω1 and ω2 are crucial to the range in which the control strategy operates and the final performance of the composite observer, making their selection particularly important. Determining the switching points empirically cannot fully utilize the advantages of the two algorithms within their respective intervals, thus affecting the observer's performance. The sliding mode algorithm estimates rotor information using back EMF, which is not a primary consideration when designing the transition interval; therefore, the high-frequency injection method is used to calculate the switching points.

[0028] Step 5: Selection of the weighting function for the composite observer. After merging the phase-locked loops of the high-frequency injection method and the sliding mode algorithm, the weighting coefficients of the error signal are determined using the speed switching point calculated in Step 4. When the speed is below the lower limit switching point ω1, the position error signal is entirely provided by the difference between the high-frequency injection method and the composite observer. If the speed is above the upper limit switching point ω2, the error is entirely derived from the difference between the sliding mode algorithm and the composite observer. When the speed is within the transition range, the error is composed of both the high-frequency injection method and the sliding mode algorithm.

[0029] Furthermore, the specific process of deriving the mathematical model of the motor under high-frequency excitation in step 1 is as follows:

[0030] The rotor voltage drop of a synchronous reluctance motor is negligible at low speeds and with large injected voltage amplitudes, so the voltage equation under high-frequency excitation is expressed as follows:

[0031]

[0032] In the formula u din and u qin i represents the high-frequency component of the stator voltage of the synchronous reluctance motor in the dq axis system. din and i qin For the high-frequency components of the stator current in the same shaft system, L d L is the direct-axis inductance of the motor. q This is the quadrature axis inductance of the motor.

[0033] In the rotor shaft system dq, the stator inductance of the motor is expressed as:

[0034]

[0035] The inductance matrix in the stationary coordinate system can be derived from equation (2):

[0036]

[0037] Where L αβ Given the inductance value in a two-phase rotating coordinate system, equation (3) shows that the inductance matrix contains rotor position information θ. e Then the observation axis system The relationship between medium and high frequency voltage and current is as follows:

[0038]

[0039] In the formula and To estimate the synchronous rotating coordinate system of the rotor High-frequency components of the voltage. and For the high-frequency component of the current; define L = (L d +L q ) / 2 is the average inductance, ΔL=(L q -L d If ) / 2 is the half-differential inductance, then equation (4) can be simplified to:

[0040]

[0041] Furthermore, the specific process for selecting the high-frequency signal injection strategy in step 2, the transition interval, is as follows:

[0042] In the zero-low speed stage, high-frequency injection is used to obtain rotor information, which requires injecting a high-frequency square wave signal with an amplitude of one-tenth of the DC bus voltage into the system. High-frequency voltage introduces noise and torque ripple, affecting the estimation accuracy of the sliding mode observer in the mid-to-high speed range. Therefore, the injection range of the high-frequency signal should not exceed the upper limit switching point.

[0043] Immediately cutting off the high-frequency signal when the rotational speed reaches the upper limit switching point will cause a voltage surge, resulting in system oscillation. If injection continues, the high-frequency signal will disturb the sliding mode algorithm, exacerbating system errors. Therefore, when the rotational speed leaves the transition range and enters the medium-to-high speed stage, the high-frequency signal is smoothly cut off with linear amplitude decay.

[0044] The amplitude variation function of a high-frequency signal is expressed as follows:

[0045]

[0046] in To estimate the rotational speed for the composite observer, U in To inject the high-frequency signal amplitude, ω2 is the upper limit switching point of the rotational speed, and ω3 is the rotational speed when the high-frequency signal decays to zero.

[0047] When estimating the rotational speed When the speed is less than ω2, the high-frequency signal amplitude remains unchanged; when the estimated rotational speed is between ω2 and ω3, the signal amplitude decreases linearly with increasing rotational speed; once the rotational speed exceeds ω3, the high-frequency signal is completely cut off. Compared to immediately stopping the injection, the linearly decaying high-frequency signal switching strategy can effectively reduce system losses and improve the observer accuracy.

[0048] Furthermore, the specific process of designing the composite position observer based on the fused phase-locked loop in step 3 is as follows:

[0049] The injected high-frequency signal causes fluctuations in the sliding mode algorithm's estimation of the rotor position angle. Angular velocity is calculated from the rate of change of angle between two adjacent sampling periods. Fluctuations in the position angle cause severe oscillations in the velocity amplitude, thus affecting the accurate estimation results of the high-frequency injection method in the zero-low speed range. This results in poor performance of composite observation using angles. Therefore, a completely new composite observer needs to be built using the position error as the phase-locked loop tracking signal.

[0050] First, the phase-locked loops of the high-frequency injection method and the sliding mode algorithm are combined. The estimated position information of the two is subtracted from the final position of the composite observer and then fed back to the phase-locked loop. The position errors of the two control strategies are weighted and fused within the transition interval, and the error weights are adjusted according to the running speed to change the amount of error fed back to the phase-locked loop.

[0051] The core idea of ​​this composite position observer is to abandon the use of angle signals and instead smooth the error signals.

[0052] Furthermore, the specific process for calculating the transition interval switching point in step 4 is as follows:

[0053] The switching points ω1 and ω2 are crucial in determining the range within which the control strategy operates and the final performance of the composite observer. Determining the switching points empirically fails to fully utilize the advantages of each algorithm within its respective range, thus affecting the observer's performance. The sliding mode algorithm estimates rotor information using back EMF and is not a primary consideration when designing the transition range. The derivation of the switching point using the high-frequency injection method is as follows.

[0054] Simplified equations for high-frequency current in estimating synchronous rotation In a coordinate system, it can be represented as:

[0055]

[0056] in The d-axis high-frequency current component. U is the q-axis high-frequency current component. in Let ω be the amplitude of the injected high-frequency signal. in R is the frequency of the injected high-frequency signal. s Motor stator resistance, For the rotor position error signal, algebraic Algebra b = R s ω in (L d +L q ), for rotor position error signal function Linearization yields:

[0057]

[0058] Where η is the gain coefficient, if the effect of back electromotive force on the low-speed operation of the motor is not ignored, the above formula can be expressed as:

[0059]

[0060] in Substituting a and b into equation (9), we get:

[0061]

[0062] The frequency of the injected high-frequency signal is highly separated from the fundamental frequency, and the high-frequency impedance is numerically much greater than the stator resistance. Therefore, equation (10) can be approximately equivalent to... Re-deriving equation (9) yields:

[0063]

[0064] The transition interval switching point can be determined using the system position error signal, with the error interval set as (m, n) and the speed switching interval as (ω1, ω2). If the position error between the high-frequency injection method and the composite observer reaches the lower limit of the error range, the weighting function will readjust the estimated values ​​of the two control strategies. Similarly, when the error between the sliding mode algorithm and the composite observer reaches the upper limit, the weighting function will also be re-estimated. Therefore, the formula for calculating the transition interval switching point can be obtained:

[0065]

[0066] Where m is the lower limit of the error interval and n is the upper limit of the error interval. Substituting the upper and lower limits of the position error of the composite observer as 0.08 rad and 0.12 rad respectively into the above formula, the speed switching points ω1 = 60 r / min and ω2 = 100 r / min can be obtained.

[0067] Furthermore, the specific process for selecting the composite observer weight function in step 5 is as follows:

[0068] After merging the phase-locked loops of the high-frequency injection method and the sliding mode algorithm, the weighting coefficients of the error signal are determined using the speed switching point calculated in step 4. When the speed is less than the lower limit switching point ω1, the position error signal is entirely provided by the difference between the high-frequency injection method and the composite observer. If the speed is higher than the upper limit switching point ω2, the error is entirely derived from the difference between the sliding mode algorithm and the composite observer. When the speed is within the transition range, the error is composed of both the high-frequency injection method and the sliding mode algorithm, and the weighting function expression is as follows:

[0069]

[0070] in: The rotor position estimate is based on the improved high-frequency injection method. The estimated value is based on the improved sliding mode algorithm. This is the rotor electric angular velocity output by the composite observer, and the weighting coefficient λ can be defined as:

[0071]

[0072] Example

[0073] The sensorless control system for a synchronous reluctance motor based on a fused phase-locked loop is designed as follows: AC input is a 335V three-phase AC power grid; the motor has 2 pole pairs; rated speed is 1500 r / min; rated torque is 9.5 N·m; stator resistance is 2 Ω; and moment of inertia is 4.2 × 10⁻⁶ N·m. -4 kg·m 2 The system has a direct-axis inductance of 65mH, a quadrature-axis inductance of 33mH, and a switching frequency of 10kHz.

[0074] To verify the superiority of the sensorless control system based on fused phase-locked loop in suppressing system chattering and resisting load disturbances, as well as the accuracy of motor speed and rotor position estimation, MATLAB simulation was used to compare the output speed and rotor position information of the system under two transient algorithm control methods, while ensuring that all system parameters are consistent.

[0075] Figure 3 This is a block diagram of a sensorless composite control structure for a synchronous reluctance motor across its entire speed range.

[0076] Figure 4 This is a speed response curve of the improved composite position observer based on a fused phase-locked loop. The graph shows that when the speed passes the upper and lower limit switching points and the sensorless control strategy changes, the speed fluctuation amplitude decreases compared to the previous version, reaching approximately 12 r / min.

[0077] Figure 5 This is a comparison chart of rotor positions using an improved composite position observer based on a fused phase-locked loop. The chart shows that the improved composite observer effectively alleviates position tracking delay during the transition phase.

[0078] Therefore, the comparison of the simulation waveforms above shows that the sensorless control system based on the fusion phase-locked loop is significantly better than the traditional transient algorithm control system in suppressing system chattering and resisting load disturbances, and also improves the accuracy of output speed and rotor position information.

[0079] The above discussion is merely one embodiment of the present invention. Any equivalent modifications made based on the present invention are included within the scope of patent protection of the present invention.

Claims

1. A sensorless control method for the entire speed range of a synchronous reluctance motor, characterized in that, Includes the following steps: Step 1: Construct a mathematical model of a synchronous reluctance motor under high-frequency excitation; Step 2: Select the high-frequency signal injection strategy in the transition range. When the motor speed is in the zero low speed range, the high-frequency injection method is used to obtain rotor information. At this time, a high-frequency square wave signal with an amplitude of one-tenth of the DC bus voltage needs to be injected into the system. The range of the injected high-frequency signal should not exceed the upper speed limit switching point. Step 3, design of composite position observer based on fused phase-locked loop; first, the phase-locked loop of high frequency injection method and sliding mode algorithm is merged, the position estimation information of the two is subtracted from the final position of composite observer and fed back to phase-locked loop, the position errors of the two control strategies are weighted and fused in the transition interval, and the error weight is adjusted according to the motor running speed to change the amount of error fed back to phase-locked loop; Step 4: Calculate the transition interval switching point; determine the transition interval switching point using the system position error signal, and set the error interval as... Speed ​​switching range If the positional error between the high-frequency injection method and the composite observer reaches the lower limit of the error range, the weighting function will readjust the estimated values ​​of the two control strategies; when the error between the sliding mode algorithm and the composite observer reaches the upper limit, the weighting function will also re-estimate; thus, the formula for calculating the transition interval switching point can be obtained: (12) in This is the lower limit of the error interval. This represents the upper limit of the error range; For average inductance, It is a half-differential inductor. For the direct-axis inductance of the motor, For the quadrature axis inductance of the motor, For the stator resistance of the motor, The frequency of the injected high-frequency signal; Step 5, selecting the weighting function of the composite observer; after merging the phase-locked loops of the high-frequency injection method and the sliding mode algorithm, the weighting coefficients of the error signal are determined using the speed switching point calculated in Step 4; when the speed is less than the lower limit switching point, the position error signal is entirely provided by the difference between the high-frequency injection method and the composite observer; if the speed is higher than the upper limit switching point, the error is entirely from the difference between the sliding mode algorithm and the composite observer; when the speed is in the transition range, the error is composed of both the high-frequency injection method and the sliding mode algorithm.

2. The sensorless control method for the full speed range of a synchronous reluctance motor according to claim 1, characterized in that, The specific derivation process of the mathematical model of the synchronous reluctance motor under high-frequency excitation in step 1 is as follows: The voltage equation under high-frequency excitation is expressed as follows: (1) In the formula and for High-frequency components of the stator voltage of a synchronous reluctance motor in the shaft system. and These are the high-frequency components of the stator current under the same shaft system. In the rotor shaft system In this context, the stator inductance of the motor is represented as: (2) The inductance matrix in the stationary coordinate system can be derived from equation (2): (3) in The inductance value is given in a two-phase rotating coordinate system. Equation (3) shows that the inductance matrix contains rotor position information. Then the observation axis system The relationship between medium and high frequency voltage and current is as follows: (4) In the formula and To estimate the synchronous rotating coordinate system of the rotor High-frequency components of the voltage. and For the high-frequency component of the current; Equation (4) is simplified to: (5)。 3. The sensorless control method for the full speed range of a synchronous reluctance motor according to claim 1, characterized in that, The specific process for selecting the high-frequency signal injection strategy in step 2 is as follows: To obtain rotor information during the zero-low speed stage using the high-frequency injection method, a high-frequency square wave signal with an amplitude of one-tenth of the DC bus voltage needs to be injected into the system; the injection range of the high-frequency signal should not exceed the upper limit switching point. The amplitude variation function of a high-frequency signal is expressed as follows: (6) in To estimate the rotational speed for the composite observer, To inject high-frequency signal amplitude, This is the point where the maximum speed limit is switched. The rotational speed when the high-frequency signal attenuates to zero; When estimating the rotational speed Less than At that time, the amplitude of the high-frequency signal remains unchanged; the estimated rotational speed is located at... and Between these two points, the signal amplitude decreases linearly with increasing rotational speed; once the rotational speed exceeds a certain threshold... High-frequency signals are completely removed.

4. The sensorless control method for the full speed range of a synchronous reluctance motor according to claim 1, characterized in that, The specific process of designing the composite position observer based on the fused phase-locked loop in step 3 is as follows: First, the phase-locked loops of the high-frequency injection method and the sliding mode algorithm are combined. The estimated position information of the two is subtracted from the final position of the composite observer and then fed back to the phase-locked loop. The position errors of the two control strategies are weighted and fused within the transition interval, and the error weights are adjusted according to the running speed to change the amount of error fed back to the phase-locked loop.

5. A sensorless control method for the full speed range of a synchronous reluctance motor according to claim 1, characterized in that, The specific process for calculating the transition interval switching point in step 4 is as follows: Simplified equations for high-frequency current in estimating synchronous rotation In a coordinate system, it can be represented as: (7) in The d-axis high-frequency current component. This represents the q-axis high-frequency current component. The amplitude of the injected high-frequency signal, For the rotor position error signal, algebraic Algebra For rotor position error signal function Linearization yields: (8) in Let be the gain coefficient. If the effect of back electromotive force on the motor during low-speed operation is not ignored, the above formula can be expressed as: (9) in The arctangent coefficient, which contains rotor information, is expressed as... ,Will Substituting into equation (9), we get: (10) Equation (10) can be approximately equivalent to: By re-deriving equation (9), we can obtain: (11)。 6. A sensorless control method for the full speed range of a synchronous reluctance motor according to claim 5, characterized in that, The position error limits of the composite observer are set to 0.08 rad and 0.12 rad, respectively. Substituting these values ​​into equation (12) yields the rotational speed switching point. =60r / min, =100r / min.

7. The sensorless control method for the full speed range of a synchronous reluctance motor according to claim 1, characterized in that, The specific process for selecting the composite observer weighting function in step 5 is as follows: After merging the high-frequency injection method with the phase-locked loop of the sliding mode algorithm, the weighting coefficients of the error signal are determined using the speed switching point calculated in step 4; when the speed is less than the lower limit switching point... At that time, the position error signal is entirely provided by the difference between the high-frequency injection method and the composite observer; if the rotational speed is higher than the upper limit switching point The error comes entirely from the difference between the sliding mode algorithm and the composite observer; when the rotational speed is within the transition range, the error is composed of both the high-frequency injection method and the sliding mode algorithm, and the weighting function expression is as follows: (13) in: The rotor position estimate is based on the improved high-frequency injection method. The estimated value is based on the improved sliding mode algorithm. This is the rotor electric angular velocity output by the composite observer, with weighting coefficients. Defined as: (14)。 8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method as described in any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method as described in any one of claims 1-7.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the method described in any one of claims 1-7.

Images