本論文利用智慧型控制理論中高斯小腦模型技術來設計控制器,並植入切換式磁阻馬達驅動系統中,所設計的控制器使系統速度響應具有高性能的表現。切換式磁阻馬達相當適合應用於變轉速驅動的場合,具有高功率密度、高轉矩、高效率、無轉子繞組以及低成本等優點。然而,切換式磁阻馬達與生俱來的非線性特性,導致高性能控制器設計困難,而高斯小腦模型控制器則提供克服系統非線性特性的能力。因此本研究將切換式磁阻馬達驅動系統結合高斯小腦模型控制技術,並輔以適應性控制理論中的投影演算法(projection algorithm),設計一具線上調適參數能力的PI控制器,稱之為高斯小腦模型比例積分控制器(Gaussian Cerebellar Model Articulation PI Controller, GCMAPIC)。 為了驗證所設計控制器之性能及可行性,本研究利用dSPACE-DS1104訊號處理平台來實現所提出的控制策略。經由實驗結果證明,本研究所提出之控制策略確實可有效提升系統之動態響應。
This thesis adopts the technique of Gaussian cerebellar model in intelligent control to design the speed controller for the switched reluctance motors (SRMs) drive system. The designed system acquires superior performance in speed response. SRMs are known suitable for variable speed drive applications due to their high power density, large torque, high efficiency, no rotor windings and low cost. However, the inherent nonlinear characteristics make SRM difficult to control. Since the Gaussian cerebellar model technique provides a good capability to deal with nonlinear characteristics, this thesis proposes a Gaussian cerebellar model based online self-tuning PI controller, named Gaussian Cerebellar Model Articulation PI Controller (GCMAPIC), with the projection algorithm used in the adaptive control theory to tune the parameters. To verify the feasibility and practicality of the controller, a dSPACE-DS1104 platform is used to implement the proposed control scheme. From the experimental results, it is seen that the dynamic performance of the SRM driver system is improved by the proposed scheme.