Design and optimisation of the performance of switched reluctance motor drives.

摘要

Switched reluctance machine (SRM) has been gaining attentions from researchers and industries decades ago. Nonlinearities of SRM complicate simulation. SRM also output high torque ripples, which limit the applications of SRM. This thesis concentrates on the studies in respect to simulation model, online torque estimation, instantaneous torque control and position control.;A simulation model with nonlinearities is presented. Since it is developed under MATLAB/SimPowerSystems, it can simulate SRM, power drive and its associated control algorithm. With user defined data, the model considers both self and mutual coupling, and the mechanical parameters. The power drive and control algorithm can be altered using graphical interface. Simulation results show that the model is accurate and effective, as the results of the developed model are validated experimentally.;To estimate the electromagnetic torque online, a torque estimator for SRM under hysteresis current control is proposed. The co-energy is estimated based on a few machine parameters. The co-energy, rotor position and phase current are stored in the memory. In the next switching cycle, the newly estimated co-energy and the stored values are used to estimate the torque. Results demonstrate that the outputs of the estimator are similar to those of cubic-spline model.;For torque ripple reduction, a torque controller based on co-energy control is developed. A co-energy profile is firstly deduced from low-current magnetic characteristics. Since co-energy is proportional to the electromagnetic torque in both linear and nonlinear regions, the required co-energy can be calculated from the profile and torque command. The controller then regulates the co-energy accordingly. Simulation and experiment confirm that the torque ripple can be reduced significantly. Additionally, the controller requires less pre-measured data, comparing with traditional schemes. The torque controller is then extended to four-quadrant operation, to reduce torque ripple in both motoring and generating modes.;An algorithm for rotor position control is then developed. The position and torque control loops are connected in cascade. Utilising the principles of freedom internal model control, the parameters of the controllers are expressed in terms of machine parameters and required system response time. The simulation and laboratory outputs ascertain the robustness of the algorithm.