智慧型類神經滑動模式之車輛主動懸吊系統控制

Intelligent Sliding Mode Neural Network Control for Active Vehicle Suspension Systems

10.6989/JN.200812.0289

蒙以正(Yi-Jen Mon)

智慧型類神經網路 ； 滑動模式 ； 車輛主動懸吊系統 ； intelligent neural networks ； sliding mode ； vehicle active suspension systems

南亞學報

28期（2008 / 12 / 01）

289 - 301

繁體中文

汽車電子的研究近期受到很大的重視，其中舒適性方面的懸吊系統研究是其中之一。關鍵性技術在於車輛中的電子控制單元（ECU），其微控器會接收輪胎傳來的路面感測信號，進行控制量的運算，補償結果再傳給控制器，使車輛震動的幅度予以改善。本論文將利用MATLAB建立主動懸吊系統之數學模型與道路模型，其數學模型為二自由度的非線性耦合系統，道路模型為類似弦波變化之干擾模型，本論文除將進行該方法之驗證外，並進行全新的智慧型類神經演算法之設計發展。以此智慧型演算控制法則與滑動模式控制的方法為基礎，結合模糊類神經網路與適應控制的優點，探討改善其效能的方法，並應用於主動式車輛懸吊系統控制設計上，並驗證發展方法的性能與強健性。

The researches of automotive electronics get a lot of attentions in the near future, the comfort aspects of the suspension system is one of them. The major unit is the electronic control unit (ECU). The micro controller will receive the signal of tire sensor from the road surface to get the control gain to compensate the real controller so that the vibration of vehicle can be reduced.This paper will establish mathematical model and road model of vehicle suspension by using the MATLAB software. The mathematical model is a two degree of freedom coupled nonlinear system. The road model is a disturbance model like sine wave changes. In addition to the method of verification, a new intelligent sliding mode neural network design algorithm will be developed. Based on this intelligent algorithm, the sliding mode control approach will be combined to improve the performance. The developed controller of active suspension will be verified to have good performance and to achieve robustness.