Real-time Feedforward Controller Tuning Based on Estimated Response Iterative Tuning
摘要
In many practical applications, time-varying systems make it difficult to obtain accurate mathematical models. In such cases, it is desirable to design controllers that can adapt to changing environments directly from data. This paper proposes a novel data-driven real-time tuning method for the feedforward controller in two-degree-of-freedom control systems. The method extends the estimated response iterative tuning (ERIT) approach, originally developed for offline tuning, to a real-time framework. We also prove the stability of the proposed control scheme using the key technical lemma and confirm its effectiveness through experimental validation.