使用FPGA技術實現智慧型控制器於直流對直流電源轉換器

Abstract

本論文提出一個具有自我學習能力之模糊控制方法，整個控制法則包含一模糊規則調節器和一模糊控制器兩大部份，前者主要針對即時的輸出訊號誤差以及模糊推論運算決定規則調節量；後者主要將依據前者規則調節器產生的模糊規則來進行模糊運算求得所需之控制量。此方法的優點在於同時結合了即時線上學習的能力和模糊邏輯的特性，對於輸入變動或負載變動較大的受控系統有較佳的控制能力以及適應能力。最後，我們以FPGA實現此控制法則並運用在直流對直流電源轉換器上，經由實驗結果證明所提出之方法優於傳統PI控制器以及模糊控制器。

In this thesis, a control strategy with the ability of self-learning, automatically-tuned fuzzy control is proposed. The proposed control strategy is composed of two parts—rule modifier and fuzzy controller. The rule modifier is designed to compute the modification of rules according to output error and fuzzy inference operation; the fuzzy controller is designed to decide the control effort of the modified fuzzy rules. The advantage of the proposed method is that it combines on-line real-time information and fuzzy control, so it achieves satisfactory control performance and has adaptability to large input variation and load variation of controlled plant. Finally, the control algorithm is implemented via FPGA. The experimental results show that the proposed control strategy can achieve better performance than PI control and fuzzy control do.