Research On Fault Diagnosis And Fault Tolerant Control Strategies Of Hall Sensors AC Servo System

Permanent brushless DC motor(BLDCM)fed by square-wave current requires six Hall position points in an electrical cycle.Permanent magnet synchronous motor fed by sinusoidal-current requires rotor position information with high-resolution from Hall-effect sensors.High-resolution position and velocity estimation of PMAC with Hall-effect sensors and its low-cost vector control drives have been a heated discussion in the research area in recent years.However,in most proposed methods,the fault states of the Hall-effect sensors signals are seldom considered,and few research studies on fault diagnostic and compensations methods are able to overcome the failures of Hall-effect sensors.Therefore,the fault diagnosis and fault-tolerant control technology of Hall position sensor in ac servo drive system has very important practical significance.Firstly,this paper analyzes the influence of several common PWM modulated modes on motor commutation torque ripple.MATLAB/simulink simulation model of four PWM modulated modes was established by means of logical digital circuit modeling.By means of theoretical analysis,simulation waveform and experimental result,the influence of modulation method on torque ripple is verified.In this paper,the influence of pwm-on-pwm simulation model on the non-commutation torque ripple is studied.Secondly,aiming at the permanent magnet brushless motor driven by Hall position sensor,the fault type diagnosis method of Hall position sensor is studied.In this paper,the fault types of Hall position sensors are divided into two major categories: 1)deviations of electrical and mechanical non-uniformity associated with Hall sensors and uneven magnetization of Hall position sensor poles with rotor poles;2)single and double fault caused by failure of Hall sensor function.Aiming at different fault types,a fault diagnosis method based on Hall position signal is proposed,which can detect Hall faults type accurately and quickly without depending on the hardware detection circuit.In view of the mechanical deviation of the installation position of Hall sensor,an average Hall position signal filtering method is proposed to compensate the deviation problem.This method uses software interrupt service subroutine to capture the time interval of Hall sector and to perform cumulative filter processing,then output compensation simulation Hall position signal.The experimental results show that the proposed method not only has good steady-state performance,but also compensates for the dynamic process such as acceleration or deceleration of motor.The fault tolerant control strategy of Hall position sensor(single and double fault)is proposed for Hall sensor failure.In particular,the rising or falling edges of Hall signal transitions arrival,Hall sensors fault types are classified according to the specific type of Hall signal transitions and the current state of Hall signal;Based on the analysis of the mechanical motion equation of PMAC motor,three kinds of improved fault-tolerant strategies are proposed,1)the Linear extrapolation algorithm 2)the First-Order algorithm 3)reduced-order observer.The Hall sensor fault tolerant control strategy is tested on the ac servo motor experiment platform fed by the square and sine wave current,and the performance of the different fault tolerant control strategy of Hall sensors is verified.Experiments show that based on the reduced order observer of Hall sensor fault tolerant control technology has better fault-tolerant performance than the other two methods,which has been devised to be free from the mechanical parameters(such as,inertia)in rotor position estimation process.Finally,the effectiveness of the proposed method is verified by extensive theoretical analysis and sufficient experimental results.