Research On Control Strategies Of Unipolar Sinusoidal Excited Switched Reluctance Motors

Unipolar sinusoidal excitation,as an emerging excitation mode,can effectively suppress vibration and noise for Switched Reluctance Motors(SRMs).However,such excitation mode still suffers from several emerging challenges,such as the inability to achieve four-quadrant operation in SRMs,the need for improved accuracy in current control on the dq0 reference frame,long adjustment times in the speed control system,and the presence of torque ripple.To systematically address these issues,the key contributions of this dissertation are as follows:Initially,the unipolar sinusoidal excitation in Switched Reluctance Motors(SRMs)is constrained by the zero-sequence current reference methods,hindering its ability to attain four-quadrant operation.A novel zero-sequence current reference method for unipolar sinusoidal excitation is proposed,which overcomes the constraints,enabling four-quadrant operation and expanding the application scenarios of SRMs under unipolar sinusoidal excitation.To enhance the control performance of SRMs under four-quadrant operation,a high-precision analytical model is established in the dq0 reference frame.Furthermore,to address the unknown self-inductance coefficient in the model,an efficient model parameter identification method is proposed,combining the torque-balanced method and Fourier series fitting.The proposed identification method is both practical and effective,achieving accurate identification of the self-inductance coefficient.Experimental results demonstrate that the novel zero-sequence current reference method successfully realizes four-quadrant operation.Moreover,the model incorporating a second-order self-inductance coefficient exhibits high accuracy,providing an accurate description of the operational characteristics of SRMs under unipolar sinusoidal excitation.Secondly,for the current control of the unipolar sinusoidal excited SRMs,some challenges,such as control output delays and uncertainties in model parameters,lead to a decline in the effectiveness of dq0-axis current control.To address these issues,a discrete current prediction model is established,and an adaptive deadbeat predictive current control with characteristics of delay compensation and disturbance suppression is proposed.By enhancing the prediction mechanism of the deadbeat control algorithm and predicting the reference voltage for the next control cycle,the problem of control output delays is effectively overcome by the proposed method.To mitigate uncertainties in model parameter,an adaptive disturbance observer is designed based on the current prediction model,with consideration of delay compensation.Furthermore,the range of adaptive gains that ensure system stability is proved,effectively suppressing the impact of parameter perturbations on current control,and enhancing the robustness and current tracking performance.On this basis,a three-dimensional space vector modulation method for an asymmetrical half-bridge inverter is proposed,further reducing total current harmonic distortion.Simulation studies and experimental results validate the effectiveness of the proposed adaptive deadbeat predictive current control algorithm and the three-dimensional space vector modulation method.These approaches prove to be successful in significantly improving the accuracy of dq0-axis current tracking and reducing harmonic distortion.Thirdly,due to the presence of model parameter perturbations and external load disturbances,the performance of the SRM speed control system degrades.To enhance the speed control performance of SRMs under unipolar sinusoidal excitation,a predictive speed control strategy based on the strong tracking filter is proposed in this dissertation.By designing a strong tracking filter with multiple optimal fading factors,rapid overall approximation of model parameter perturbations and external load disturbances is achieved in the prediction model,improving the dynamic response performance of the speed control system.Through adjusting the controller inputs and constructing a new speed prediction model,the oscillations in the control law are effectively suppressed when the system reaches the desired speed,enhancing the steady-state performance of the SRM speed control system under unipolar sinusoidal excitation.Simulation and experimental results validate that the strong disturbance suppression performance for parameter perturbations and external load disturbances of the proposed strong tracking filter-based predictive speed control algorithm.Ultimately,due to mismatch of the inverse torque model,the existing torque ripple suppression strategy for unipolar sinusoidal excited SRMs experiences decreased control performance.A variable i_q torque ripple suppression strategy with a model correction mechanism in proposed in this dissertation.A high-precision torque inverse model is established based on iterative learning control and a recursive least squares estimator in this strategy,achieving high-precision real-time conversion from torque to current and accurate estimation of model parameter changes.By accurately estimating and online correcting the changes of the inverse model parameters,the torque inverse model mismatch problem is effectively overcome.The proposed torque ripple suppression strategy maintains strong capabilities for torque ripple reduction even in situations of poor inverse model matching.Simulation studies and experimental results demonstrate that the proposed variable i_q-type torque ripple suppression strategy efficiently addresses the inverse model mismatch problem,effectively suppressing torque ripple for unipolar sinusoidal excited SRMs.