半車主動式懸吊系統之最佳模糊控制器設計

Abstract

本論文利用以Takagi-Sugeno (T-S)模型為基礎的最佳模糊控制器來設計一個非線性半車主動式懸吊系統。我們有兩個方法獲得非線性懸吊系統的T-S模糊模型。第一種方法是從半車懸吊系統的數學模型直接推導，另一種方法由一個六層的修正自我建構模糊類神經推論網路來得到懸吊系統的T-S模型。有了懸吊系統的輸出和輸入資料，修正自我建構模糊類神經推論網路可以動態地增加模糊規則，使模型輸出誤差最小化。位於車體和車輪之間的懸吊系統會經由車體的垂直位移和前傾位移產生垂直力，所以路面不規則性是影響乘客舒適性的主要因子，一個主動式懸吊系統的控制器必須能夠減少垂直和前傾位移量。所推導出的最佳化模糊控制器經由模擬顯示其可行性及實用性，而且比一般的被動式懸吊系統的性能還好。

The optimal fuzzy controller of a nonlinear active suspension system is designed based on the Takagi-Sugeno (T-S) fuzzy model. There are two ways to obtain the T-S fuzzy model. First, we can convert the mathematical model of the half-car suspension system into a T-S fuzzy model directly. Second, we can use a six layer modified self-constructing neural fuzzy inference network (modified SONFIN) to generate its T-S fuzzy model. With proper input and output data of the half-car suspension system for training, the modified SONFIN can dynamically increase the number of fuzzy rules, and meanwhile minimize the model output error. The vertical displacement of suspension elements between the vehicle body and the wheels will generate vertical forces which will excite both heave and pitch motions, and road irregularities are the main factor affecting ride comfort of passengers. An active suspension controller must be able to minimize both heave and pitch movements. Simulation results indicate the feasibility and the applicability of the designed controller. Results show that the developed controller provides better performance than conventional passive suspension system.