本論文提出一個切換式PD-PI控制器(Switching-Type PD-PI Controller, SWPD-PI)，此控制結構整合PD可以改善暫態響應與PI控制器可以消除穩態誤差的特性來讓受控系統具有不錯的整體控制性能。在控制參數的選取上，本論文應用基因演算法(Genetic Algorithm, GA)與粒子群最佳化(Particle Swarm Optimization, PSO)演算法等兩種最佳化演算法來選取適當的切換值以及PD與PI的控制參數。在適應函數(fitness function)的選取上，本論文提出一個新的評估系統性能的方式，其使得最佳化演算法更可以找到讓系統整體性能更好之控制參數。最後，本論文將所提的方法應用於自動電壓調整器系統(Automatic Voltage Regulator, AVR)上，從一些結果的比較中可以證實所提方法確實可以讓受控系統具有不錯的整體控制性能。 In this thesis, a switching-type PD-PI controller and a parameter selection method are proposed. The proposed control structure combines the features that PD controller can improve the transient response and PI controller can erase the steady error to let the controlled system with a good performance in the global response. In the parameters selection, two optimal algorithms of Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) algorithm are applied to select an appropriate switching value and control parameters in PD and PI controllers so that the controlled system has a good performance. In the selection of the fitness function, a new evaluation method in the system performance is proposed so that GA or PSO algorithm can effectively find a parameter set which lets the controlled system have a better performance. Finally, the proposed method is applied in the automatic voltage regulator. From some simulation and comparison results, we can find that the switching-type PD-PI control structure and the proposed parameter selection method can effectively find a parameter set with a better control performance than the other methods.
