Hall screening and suspension performance improvement method of bearingless motor in hall asymmetry case

Abstract

The application discloses a Hall screening and suspension performance improvement method for a bearingless motor under a Hall asymmetry, and belongs to the technical field of motor control. In view of the problem that the inherent asymmetry of a Hall leads to a decline in suspension performance, a function model with a Hall asymmetry A as an independent variable is first constructed to quantitatively describe the influence of A on displacement detection; and a closed-loop transfer function fusing A is then established. On this basis, a Hall screening strategy based on a critical proportionality degree method is proposed: the integral and differential parameters are set to zero, the proportional gain is gradually increased to obtain a critical gain K U and a critical period T U , the to-be-tested control parameters are calculated according to a PID parameter setting rule, and are substituted into the system to determine whether the displacement converges, so as to determine the availability of the Hall off-line. For the available Hall, the optimized PID control parameters are directly used, and repeated trial and error is not needed. The application can quickly screen qualified Halls and automatically set optimal PID parameters, and significantly improves the stability of a suspension system and the debugging efficiency.

Description

Technical Field

[0001] This invention belongs to the field of motor body design technology, and mainly relates to a method for Hall screening and levitation performance improvement of bearingless motors under Hall asymmetry. Background Technology

[0002] BPMSMs achieve rotor levitation by generating a stable levitation force through a one-pole difference between the permanent magnet and the levitation winding, offering advantages such as high integration, no lubrication required, and no frictional wear. Currently, BPMSMs are widely used in applications requiring high purity, such as ventricular assist devices and semiconductor cleaning pumps. As a core component of artificial heart pumps, the performance of the BPMSM directly affects the system's stability.

[0003] Traditional bearingless motor rotor displacement detection devices generally employ eddy current sensors, resulting in large system size and high cost. Linear Hall sensors can effectively solve this problem. The literature "Radial Displacement Detection Method for Bearingless Permanent Magnet Thin-Sheet Motors Based on Hall Negative Sequence Demodulation" integrates Hall sensors for dual functionality, enabling a single sensor to simultaneously identify rotation angle and displacement, significantly reducing system size and cost. In this method, the health (meaning normal operation, but with imperfectly symmetrical Hall signals) and consistency of the Hall sensors determine the motor's levitation performance. However, even in a healthy state, the inherent differences among the six Hall sensors can still cause the positive half-axis of the output signals of two opposing Hall sensors to be symmetrical while the negative half-axis deviates when detecting the same magnetic field, thus causing displacement detection errors and affecting the rotor's levitation stability.

[0004] In 2020, Yan Pengyu, Zhu Zhiyu, and others from Jiangsu University of Science and Technology published a paper titled "Fault-Tolerant Control of Permanent Magnet Synchronous Motor under Hall Sensor Fault" in the Journal of Power Electronics Technology. This paper employed a traditional average speed algorithm, built a current sliding mode observer based on rotor position estimation, extracted rotor position using a phase-locked loop, and weighted the average speed algorithm results with the sliding mode observation results for different fault conditions. However, this paper only considered the fault tolerance of angle identification under abnormal Hall signal conditions and did not study displacement. In 2024, Zhao Pan from Nanjing University of Aeronautics and Astronautics published a paper titled "Fault-Tolerant Detection of Radial Displacement of Bearingless Permanent Magnet Thin-Layer Motor under Single Hall Fault" in the Journal of Electrical Engineering Technology. This paper addressed the problem of coupling angle and radial displacement information caused by single Hall faults, employing the flux linkage reconstruction idea and using residual non-fault Hall reconstruction to calculate displacement. Simulation and experiments verified the algorithm's accuracy and improved the motor's fault tolerance capability. However, it did not study the problem of displacement identification errors caused by the inherent differences in the Hall's health state.

[0005] This self-difference mainly comes from two aspects: one is the difference in sensitivity and zero offset caused by the manufacturing process deviation of Hall elements, resulting in ex-factory differences; the other is the inevitable human error made by engineers when setting the zero point and sensitivity of Halls.

[0006] Although the above deviations are small, they can also cause increased displacement ripple and even suspension failure in severe cases. However, Halls selected by engineers are generally considered as qualified products. When it is found that the suspension performance is average or suspension cannot be achieved, engineers will spare no effort to repeatedly adjust the PID parameters of the suspension control system, which wastes a lot of time. This inefficient debugging not only prolongs the R&D cycle, but also covers up the essential root cause of the problem—the inherent asymmetry of Hall sensors, a long-neglected system-level deviation.

[0007] Existing studies have tried to optimize PID parameters by establishing the closed-loop transfer function of the suspension system and using the Routh criterion, but this method has the following problems:

[0008] (1) For Halls with a large Hall asymmetry degree A that cannot achieve suspension at all, it is impossible to directly judge their availability in one step,

[0009] (2) For Halls with a large A that cannot achieve suspension at all, they are defaulted to be available, and repeatedly trial-and-error PID parameters wastes a lot of time. Summary of the Invention

[0010] Object of the invention: To solve the above problems, the present invention provides a Hall screening and suspension performance improvement method for bearingless motors under Hall asymmetry conditions, which constructs a function model with Hall asymmetry degree A as the independent variable; establishes a closed-loop transfer function integrating the characteristics of the asymmetry function; proposes a hierarchical screening strategy for Hall elements based on the asymmetry degree A; designs a system parameter tuning method with the asymmetry degree A as the optimization variable. This method can directly judge the availability of Halls offline, and directly obtain optimized PID parameters according to the value of A for available Halls.

[0011] Technical solution: To achieve the above object, the technical scheme adopted by the present invention is:

[0012] A Hall screening and suspension performance improvement method for a bearingless motor under Hall asymmetry condition, comprising the following steps: Step 1: constructing a function model with Hall asymmetry degree A as an independent variable, and establishing a closed-loop transfer function G(s) integrating the Hall asymmetry degree A, where s is a Laplacian operator, and the Hall asymmetry degree A is used to characterize the deviation degree of Hall sensor sensitivity relative to an ideal value; Step 2: for a group of Halls to be screened, obtaining the fixed Hall asymmetry degree A thereof, substituting the A and motor parameters into the closed-loop transfer function G(s), and letting the integral control parameter K I and the differential control parameter K DAll values ​​are zero; gradually increase the proportional control parameter K. P When the system exhibits sustained constant-amplitude oscillations, the proportional control parameter at this point is recorded as the critical gain K. U The period of the constant-amplitude oscillation is recorded as the critical period T. U Step 3: Based on the critical gain K U and critical period T U Step 4: If the displacement does not converge, the Hall corresponding to the current Hall asymmetry A is determined to be unusable; if the displacement converges, the Hall corresponding to the current Hall asymmetry A is determined to be usable, and the control parameters to be tested are used as the optimized PID control parameters for this set of Halls. Preferably, the specific steps in step 1 for constructing a function model with Hall asymmetry A as the independent variable include: Taking phase A Hall as an example, when the sensitivity of phase A Hall is greater than the ideal value, the output deviation of phase A Hall is recorded as follows: ,in and Let l be the displacement coefficient and l be the length of the rotor displacement vector. The rotor displacement vector angle, The electric angular frequency of the motor. The initial phase of phase A Hall; the output deviation is compared with... Multiplying and extracting the fundamental component, the influence value of the Hall asymmetry A on the x-axis displacement is obtained. .

[0013] Preferably, the closed-loop transfer function G(s) of the fused Hall asymmetry A established in step 1 is:

[0014]

[0015] in, This is the eccentric force coefficient. For rotor mass, , The angular frequency associated with the Hall signal. For the Laplace transform of the displacement command, This is the Laplace transform for displacement feedback.

[0016] Preferably, the specific method for checking whether the rotor displacement converges in step 3 is as follows: after substituting the control parameters to be checked into the closed-loop transfer function, solve its characteristic equation. If all characteristic roots are located in the left half-plane of the complex plane, the displacement is determined to converge; if there are characteristic roots located in the right half-plane, the displacement is determined not to converge.

[0017] Preferably, in the PID parameter tuning rule, the derivative time constant T D The critical period T U Determined proportionally.

[0018] The beneficial effects that can be achieved by adopting the above technical solutions conceived in this invention are as follows:

[0019] (1) A quantitative function model between Hall asymmetry A and displacement detection error was established for the first time, and A was incorporated into the closed-loop transfer function of the suspension system, so that the influence of the inherent asymmetry of Hall sensor on system stability could be quantitatively analyzed, providing a theoretical basis for Hall screening and parameter optimization.

[0020] (2) After calculating the control parameters to be tested according to the PID tuning rules based on the critical parameters obtained by the critical proportional method, the rotor displacement is re-substituted into the closed-loop transfer function to check whether it converges. If it converges, the Hall effect sensor is usable; if it does not converge, the Hall effect sensor group is directly eliminated. This method can determine the availability of the Hall effect sensor offline in one go without building an actual suspension system, avoiding the large amount of time wasted in the traditional method due to the inability to determine the availability of the Hall effect sensor in advance and the need to repeatedly try to adjust the PID parameters.

[0021] (3) For qualified Hall sensors that pass the displacement convergence test, the control parameters to be tested are directly used as the optimized PID parameters, eliminating the need for repeated manual adjustments and significantly improving the debugging efficiency and suspension stability of the suspension system.

[0022] (4) This invention combines PID parameter tuning rules with Hall asymmetry A, closed-loop transfer function and displacement convergence test to form a complete closed-loop technology of modeling-critical parameter acquisition-displacement convergence test-Hall screening-parameter output, which is different from the conventional use of PID tuning rules for parameter calculation in the prior art. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the closed-loop transfer function constructed in this invention, with the Hall asymmetry degree A as the independent variable.

[0024] Figure 2 This is a pole distribution diagram for the Hall asymmetry A of the present invention.

[0025] Figure 3 The image shows the rotor displacement waveform when the Hall effect is perfectly symmetrical.

[0026] Figure 4 The image shows the rotor displacement waveform when the Hall effect is not perfectly symmetrical.

[0027] Figure 5 It is a motor structure.

[0028] Figure 6This is the x-axis displacement waveform after PID parameter optimization when the asymmetry A=1 in this invention.

[0029] Figure 7 This is the x-axis displacement waveform after PID parameter optimization when the asymmetry A=5 according to the present invention. Detailed Implementation

[0030] The technical solution of the present invention will now be described in detail with reference to the accompanying drawings and embodiments. This embodiment employs the following... Figure 5 The present invention uses a bearingless permanent magnet synchronous motor with 6 stator teeth, 1 pair of torque poles, and 2 pairs of suspension poles as an example to illustrate the Hall screening and suspension performance improvement method proposed in this invention.

[0031] I. Constructing a functional model with Hall asymmetry A as the independent variable.

[0032] Taking Phase A Hall as an example, when the sensitivity of Phase A Hall is greater than the ideal value, the output deviation of Phase A Hall is recorded as follows: ,in, and Let A be the displacement coefficient and A be the Hall asymmetry. The length of the rotor displacement vector. The rotor displacement vector angle, The electric angular frequency of the motor. This is the initial phase of phase A Hall. The above output deviation is then compared with... Multiply and extract the influence value of Hall asymmetry A on the x-axis displacement. This function model quantitatively describes the displacement detection error caused by the Hall asymmetry A.

[0033] II. Establishing the closed-loop transfer function of the fusion asymmetry A

[0034] like Figure 1 As shown, the closed-loop transfer function schematic of this invention fully illustrates the system signal flow: the input is a displacement command signal, which is subtracted from the displacement feedback signal by an adder. The deviation signal is input to the PID controller, and the output control quantity is superimposed with the disturbance term including the Hall asymmetry A, entering the levitation force output stage. After acting on the rotor, it outputs the actual displacement signal and feeds it back to the input. Based on this, the closed-loop transfer function constructed by this invention is as follows:

[0035]

[0036] Where s is the Laplace operator, This is the eccentric force coefficient. , , These are the proportional, integral, and derivative control parameters for the suspension system. For rotor mass, , The angular frequency is related to the Hall signal. For the Laplace transform of the displacement command, This is the Laplace transform for displacement feedback.

[0037] In this embodiment, K=100; m=0.05kg;

[0038] The characteristic equation of the closed-loop function is:

[0039]

[0040] Pole analysis was performed to examine the distribution of poles as the asymmetry A changed.

[0041] When the system uses ideal PID parameters, its stability can be analyzed based on the poles of the closed-loop transfer function. When the Hall effect sensors are symmetrical, the system is in a stable state, with the poles in the left half-plane. However, when one or more of the six Hall sensors are asymmetrical due to Hall selection or inherent errors, the poles begin to shift towards the right half-plane, and the system becomes unstable. For cases with low Hall effect asymmetry, optimizing the PID parameters can still bring the suspending system poles back to the left half-plane, and the system can still achieve stability. However, for cases with high asymmetry, even if the optimized PID parameters keep the poles in the right half-plane, the system will become unstable and oscillate. From the transfer function, we know that A and K... P K D K I Both are related, meaning the magnitude of A is directly related to the PID control, but the availability of the Hall effect sensor cannot be directly determined according to the Routh-Hurwitz criterion. For an available Hall effect sensor, K cannot be optimized based on A. P K D K I .

[0042] Figure 2 The pole distribution diagram for the Hall asymmetry A of this invention is shown below: the horizontal axis is the real axis of the complex plane, the vertical axis is the imaginary axis of the complex plane, the left half-plane is the stable region, and the right half-plane is the unstable region. When A=0 (completely symmetric), all poles are located in the left half-plane; when A=1 (slightly asymmetric) and after PID optimization, the poles can be pulled back to the left half-plane; when A=5 (severely asymmetric), the poles cannot be pulled back to the left half-plane, and the system is unstable.

[0043] Figure 3 The rotor displacement waveform is shown when the Hall effect is fully symmetrical: the horizontal axis is time (0~10s), and the vertical axis is displacement (-2~0mm). The curve is smooth and oscillating, and quickly converges to a stable value. Figure 4The rotor displacement waveform is shown when the Hall effect is not perfectly symmetrical: the horizontal axis is time (0~10s), and the vertical axis is -3~3mm. The curve shows obvious periodic oscillations and cannot converge stably.

[0044] III. Hall Screening and Parameter Optimization Method Based on Asymmetry A

[0045] This embodiment proposes a hierarchical screening strategy for Hall elements based on asymmetry A and a system parameter tuning method with asymmetry A as the optimization variable. The steps are as follows:

[0046] 4-1. Calculate the asymmetry A of a fixed value for each selected Hall effect sensor, and then input A and the motor into... Let K I =K D =0, let K P Starting from zero, gradually increase the gain and observe the system's response. When the system exhibits sustained and constant-amplitude oscillations, the gain at this point is the critical gain K. U The period of this constant-amplitude oscillation is defined as the critical period T. U .

[0047] 4-2. Let K P = 0.6K U K I = 1.2K U / T U K D =K P *T D Input these control parameters into the system and check if the displacement converges. If the displacement does not converge, the Hall effect factor corresponding to this A should be discarded and the debugging process should not proceed. Where T... D The differential time constant is taken in this embodiment. .

[0048] 4-3. If the displacement converges, then the Hall corresponding to this A is usable, and the optimization parameter corresponding to this set of Halls is K. P =0.6K U K I = 1.2K U / T U K D =K P *T D .

[0049] IV. Verification of Examples

[0050] like Figure 5As shown, the method proposed in this invention can be used to determine the availability of Hall effect sensors corresponding to different degrees of symmetry A offline. For example, when A=1, after the above PID optimization, the system poles can be pulled back to the left half-plane, and the rotor displacement can converge. When A=5, the system poles cannot be pulled back to the left half-plane, and the Hall effect sensor is unqualified.

[0051] Figure 4 The waveform shows the rotor displacement when the Hall effect is asymmetrical, and the displacement exhibits oscillations.

[0052] After optimization of the discrimination method and parameters of this invention, the Hall effect with A=1 is usable and the displacement oscillation disappears, as shown below. Figure 6 As shown.

[0053] This invention utilizes Figure 5 The unavailability of Hall effect sensors with A=5 was directly determined. To verify this conclusion, the optimized parameters were introduced into the system. The displacements first converged and then gradually diverged, verifying the correctness of the Hall effect screening method presented in this paper. Figure 7 The displacement waveform after PID optimization when A=5 is given. It can be seen that the displacement waveform oscillates and cannot converge stably.

Claims

1. A method for Hall screening and levitation performance improvement of a bearingless motor under Hall asymmetry conditions, characterized in that, Includes the following steps: Step 1: Construct a function model with Hall asymmetry A as the independent variable, and establish a closed-loop transfer function G(s) that incorporates the Hall asymmetry A, where s is the Laplace operator, and the Hall asymmetry A is used to characterize the degree of deviation of the Hall sensor sensitivity from the ideal value; Step 2: For a group of Hall sensors to be screened, obtain their fixed Hall asymmetry A, substitute A and the motor parameters into the closed-loop transfer function G(s), and let the integral control parameter K I and differential control parameter K D All values ​​are zero; gradually increase the proportional control parameter K. P When the system exhibits sustained constant-amplitude oscillations, the proportional control parameter at this point is recorded as the critical gain K. U The period of the constant-amplitude oscillation is recorded as the critical period T. U Step 3: Based on the critical gain K U and critical period T U Calculate a set of control parameters to be tested according to the PID parameter tuning rules, substitute the control parameters to be tested into the closed-loop transfer function, and check whether the rotor displacement converges; Step 4: If the displacement does not converge, then determine that the Hall corresponding to the current Hall asymmetry A is unavailable; If the displacement converges, the Hall corresponding to the current Hall asymmetry A is determined to be usable, and the control parameter to be tested is used as the optimized PID control parameter for this group of Halls.

2. The method for Hall screening and levitation performance improvement of a bearingless motor under Hall asymmetry as described in claim 1, characterized in that, The specific steps in step 1 of constructing a function model with Hall asymmetry A as the independent variable include: Taking phase A Hall as an example, when the sensitivity of phase A Hall is greater than the ideal value, the output deviation of phase A Hall is recorded as follows: ,in and Let l be the displacement coefficient and l be the length of the rotor displacement vector. The rotor displacement vector angle, The electric angular frequency of the motor. The initial phase of phase A Hall; The output deviation is compared with... Multiplying and extracting the fundamental component, the influence value of the Hall asymmetry A on the x-axis displacement is obtained. .

3. The method for Hall screening and levitation performance improvement of a bearingless motor under Hall asymmetry as described in claim 1, characterized in that, The closed-loop transfer function G(s) of the fused Hall asymmetry A established in step 1 is: ； in, This is the eccentric force coefficient. For rotor mass, , The angular frequency associated with the Hall signal. For the Laplace transform of the displacement command, This is the Laplace transform for displacement feedback.

4. The method for Hall screening and levitation performance improvement of a bearingless motor under Hall asymmetry as described in claim 1, characterized in that, The specific method for checking whether the rotor displacement converges in step 3 is as follows: after substituting the control parameters to be checked into the closed-loop transfer function, solve its characteristic equation. If all characteristic roots are located in the left half-plane of the complex plane, the displacement is determined to converge; if there are characteristic roots located in the right half-plane, the displacement is determined not to converge.

5. The method for Hall screening and levitation performance improvement of a bearingless motor under Hall asymmetry as described in claim 1, characterized in that, In the PID parameter tuning rules, the derivative time constant T D The critical period T U Determined proportionally.

Images