本文提出改善最佳時問「碰一碰」控制器所引發之顫振現象之快速定位的新方法，稱之爲「以最佳軌跡弧長做回授之時間次最佳模糊控制法」。本法是以系統最佳動態軌跡弧長作爲回授變數，將最佳時間「碰一碰」控制法與傳統高增益回授控制法以模糊邏輯結合爲一，同時達成快速定位並消除一般最佳時間「碰一碰」控制法於實際系統中常產生的顫振現象。本文首先提出以最佳軌跡弧長做回授之動機與優點，再介紹以積分將狀態變數轉換爲沿最佳動態軌跡量測之狀態變數的方法。最後並利用對伺服馬達系統控制之模擬實驗際來說明此新方法之優點。
This paper addresses ”Optimal Path Estimation Method (OPEM)”, a new approach for the implementation of time optimal bang-bang control. The method is based upon feedback information obtained by estimating the distance between the present state to the switching surface defined by the optimal trajectory. A fuzzy controller is used for implementing the control logic. The method is able to eliminate the chattering phenomenon common to the conventional trajectory access and following controller. The system trajectory will eventually enter a finite region around the target point and will stay inside the region. The advantages of the proposed method over traditional Bang-Bang control are illustrated by the numerical simulations and experiments of a Servomotor system.
