一款菱形四輪可傾窄車之多體系統建模與使用QFT/H∞方法之強健雙迴路PID傾斜動作控制器設計

Abstract

本論文介紹一款著眼於解決石油短缺、污染與交通壅塞的電動菱形四輪可傾窄車，稱為IPM (Intelligent Personal Mobility)，此車具輕與窄的特性，車身具有傾斜功能。為分析一些冒險的行駛狀況，本文中利用平面多體系統方法對IPM建立模型；本論文中也推導出一可用於平面多體系統之輪胎力模型；而為了實驗地驗證此IPM多體模型，本文並提出一種可用於菱形四輪可傾窄車行進間的輪胎負載之算法(藉由利用兩個加速規與一個角度量測器)。此多體系統模型經由與實車曲行測試比較驗證，此一模型並可以分析車體各主要部件間連結作用力並可輔助控制器設計。 本文中利用結合QFT/Ｈ∞方法設計一可對抗路面干擾之IPM傾斜動作控制器，並提出一系統方法將此控制器轉換為較易實現的強健雙迴路PID控制器(robust double-loop PID controller)，而該雙迴路PID控制器被經由已驗證之多體系統模型驗證其特性，其具有對於來自路面干擾和不完全傾斜命令(完全傾斜命令可使得駕駛位置所受到重力回覆力矩與向心力翻覆力矩接近平衡)之負載干擾抵抗力，並具有對傾斜動作之高循跡強健性(high tracking robustness)。

In this dissertation, a personal electric narrow tilting vehicle (NTV) called “IPM (intelligent personal mobility)” is developed to solve problem of oil shortage, pollution and traffic jam. The vehicle has four wheels arranged in a diamond shape. It is capable of operating in vehicular tilts with less weight and width. In order to simulate hazardous driving conditions, IPM was built a model by using planar multibody system method. A planar multibody system tire model was proposed to represent ground-vehicle interaction. In order to verify the IPM multibody model experimentally, we also proposed a separate calculation method using two acceleration sensors and one angular position sensor for the purpose of obtaining ground forces. The multibody model was verified by comparing with the real IPM in designed slalom tests. It matched with the real vehicle effectively and accurately. This model can applied to analyze the joint reaction force of IPM to assist in controller design. An IPM tilting motion controller designed by using Quantitative Feedback Theory (QFT) and H∞ theory was proposed in this dissertation. For implement purpose, this controller was systematically translated into a double-loop PID controller (two loop PID control, one tilting position control loop enclosed one tilting rate control loop). The controller was verified by working with the verified IPM multibody model. After the verification, the designed controller was verified that it has high tracking robustness. It also was verified it can resist the load disturbance from lumpy roads and inadequate tilting command (Focus on driver position, if the gravity antiroll torque can balance with the centripetal force rollover torque, it is called adequate tilting command.) by working with the verified model.