Research On Permanent Magnet Spherical Motor Control System Using Modeling Technology And Intelligent Algorithms

Since the requirement of multiple degree-of-freedom(Multi-DOF)apparatus is enhanced by the development of industrial technology,Multi-DOF motors attracted extensive attention from scholars worldwide.The multi-DOF motor has wide application prospects in the field of applications such as robots,manipulators of intelligence manufacturing,satellite attitude control,and rotational station.A PMSpM has the advantages of compact structure,no mechanical transmission mechanism,flexible motion mode,and no accumulation of motion error.The research on permanent magnet spherical motors(PMSpM)is the major research direction for Multi-DOF motors.The exploratory study on the PMSpM torque modeling,the driving strategy,and the rotor attitude detection method prompts the process toward industrial application.This dissertation researches the PMSpM torque modeling,the intelligent algorithm based drive strategy and rotor attitude estimation algorithm.A PMSpM control system with a C++ upper software is developed.These researches provide scientific theoretical support for the potential industrial applications of the PMSpM.The researches also hold practical value and certain scientific significance.The innovation points in this dissertation include:Firstly,this dissertation proposes a spherical harmonics-based PMSpM fast analytic modeling method.The mid-geometrical-equivalence principle for the cubic permanent magnet(PM)is proposed.To revise the magnetic field wave deviation between the analytic model result and the experimental one,an engineering correction method is introduced for the magnetic analytic model according to the geometrical symmetry of the PM in both azimuth angle direction and polar angle direction.The PMSpM torque model is established by using Lorentz Force Law.FEM and experimental verifications prove the correctness of the proposed models.Secondly,the monocular vision target tracking based rotor attitude estimation algorithm is proposed.In the beginning,the kinematic analytic method is studied for the PMSpM with different sensors-based rotor attitude detection methods.Then,a multi-objective Kalman KCF tracking(MKKCF)algorithm is proposed to track three targets on the visual feature component which is mounted on top of the rotor output shaft.Furthermore,the real-time performance and the robustness of the proposed MKKCF algorithm are compared with different visual tracking algorithms.After obtaining the three targets’ position on the visual feature component by using the proposed MKKCF algorithm,the rotor attitude of the PMSpM is detected.The accuracy and robustness of comparisons among rotor attitude detection systems with different sensors indicate the significant advantages of the proposed rotor attitude estimation algorithm.Thirdly,the improved particle swarm optimization(ISPO)algorithm is proposed for the PMSpM driving strategy.The torque maps of the PMSpM are firstly established according to the proposed torque model.Then,the bilinear interpolation algorithm based PMSpM current optimization strategy is proposed,due to its better accuracy and real-time performance.To realize fast calculating of the PMSpM driving current when the desired torque is acquired,the improved adaptive inertia weight and adaptive learning factors are proposed based on the standard particle swarm optimization(PSO)algorithm.The robustness and the real-time performance of the PMSpM driving strategy algorithm are significantly improved.The full factorial design of experiments(DOE)method improves the real-time performance of the PMSpM driving strategy is further improved.Finally,on the basis of the research above,a PMSpM control system with the Qt 5.14/C++upper software is developed,and the proposed torque model and intelligent algorithms are verified in this control system.To realize the closed-loop control,this dissertation deduces the dynamics model of the PMSpM under study.Moreover,the PD control algorithm,the proposed driving strategy,and the rotor attitude estimation algorithm are deployed in the upper software.The simulation and experimental study of the typical PMSpM tilt motion proves the applicability and effectiveness of the proposed models and intelligent algorithms based rotor attitude estimation algorithm and PMSpM driving strategy.