應用CMAC於非線性控制問題

Abstract

CMAC類神經網路是一種可以有效的改善非線性控制系統的工具。由我們的 模擬和實驗的結果顯示，它的確可以有效的降低追蹤控制的誤差，但是它 也會造成系統的不穩定。在本論文中，首先我們將以一個簡單的倒立單擺 為例子來說明CMAC控制系統的控制特性。然後我們將利用一個數位化模型 的電腦模擬，來探討系統不穩定的現象。藉由各個系統參數定量的探討， 我們找出使系統不穩定的原因，並且提出一些改進的方法。最後我們將由 一個實驗來驗證我們模擬的各種結果。 The CMAC neural network is a practical tool for improving existing nonlinear control systems. Both simulation and the experiment results show that the CMAC can effectively reduce tracking error, but the results also show that the CMAC can destabilize a control system which is otherwise stable. In the thesis we will first use a simple simulation to show the effect of CMAC controller. Then by using a simple discrete time control system, we study the unstable phenomenon. The quantitive studies are presented to search for the cause of the instability in the CMAC control system. Based on these studies, methods are discussed to improve system stability. Experimental results on controlling a real world system is provided to support the findings in simualtions.