Research On Novel Drive System For Doubly Salient Permanent Magnet Motor

Doubly salient permanent magnet (DSPM) motor has been researched by many scholars due to incorporate the merits of both switched reluctance motor and PM brushless motor. In this dissertation, a novel driving system for DSPM motor is developed. The main topics include finite element analysis (FEA), the mechanism of torque ripples, the relationship between turn-on angle & turn-off angle and torque ripple, technology of torque control, robustness of speed loop, and fault diagnosis and fault tolerant of position signals.The static characteristics of 6/4-pole DSPM motor is given through two-dimensional FEA software Maxwell of ANSOFT corporation, including magnetic field distribution, winding inductances, winding flux linkages, reluctance torque and permanent magnet(PM) torque and so on. The magnetic field distribution and parametric variation of DSPM motor is analyzed after load through simulated real currents. The winding inductances and winding flux linkages for 6/4-pole doubly salient electro-magnet motor and 12/8-pole DSPM motor at no load is also given in order to describe the common character of nearly equal flux slope in former and back of commutation for DSPM motor. The nonlinear simulation model of driving system for DSPM motor based on MATLAB/SIMULINK is built according to the data of flux and induction calculated by Maxwell software.The cause of torque ripples is analyzed in detail from the view of commutation ripple and steady ripple, and a conclusion that commutation ripple is the main ripple for DSPM motor is given. A novel optimal control strategy of turn-on angle and turn-off angle is proposed in order to reduce commutation torque ripple. The relationship between the current ratio of commutation phase and torque ripple is analyzed, and the optimal current ratio is deduced. The validity of the novel control strategy is verified by simulation and experiment.It is concluded from the contrast analysis of reluctance torque and PM torque that PM torque is the main torque and the main ripple is PM torque ripple. The torque control strategy is proposed in order to further decrease the torque ripple and be fit for electric vehicle. And the torque observer is structured. Also the transition formula from line voltage to phase voltage is given and a second-order filter is designed for line voltage.More and More attentions have been paid to robustness study due to magnetic field saturation and winding resistance variation after long run, which will bring precision and performance of control system low. The sliding model control (SMC) with strong robustness is employed to decrease the influence of parameters variation. State equations are established and the relationship between the output of SMC and torque is deduced. At the same time first-order filter is employed to reduce the chattering, but the dynamic characteristic of system will be weakened. Thus a speed regulator with good dynamic characteristic is designed based on novel SMC combined feed forward control.Position sensors are the key of exact commutation for brushless motor, hence it is very necessary to study fault diagnosis and fault tolerant control of position signals. The running status after fault is described, including current, torque, level of position signal. A method of fault diagnosis through detecting the level value of three phase's position signals is proposed. Also the method of fault tolerant after one-phase fault and two-phase fault is proposed through revealing the correlation of three phase signals. Thus the reliability of system is increased and the field of application is broadened.Based on above theory analysis, a digital experimental test-bed on DSP and CPLD for DSPM motor driving system is built.