線控化底盤系統故障偵測與隔離方法與XiL模擬驗證

Abstract

本篇論文探討自主駕駛車輛之車輛動態、動力、轉向、煞車次系統的動態行為，以開發線控化底盤系統之故障偵測與隔離方法在自駕車運行時偵測次系統故障以及車輛不良行為，在必要的時候可以透過此隔離方法辨別故障來源，在意外發生前修正決策層命令。本論文研究之車輛是一三噸重電動貨卡其動力系統搭載95kw的永磁同步馬達作後輪驅動，轉向系統與煞車系統皆已線控化由伺服馬達取代踏板與方向盤作驅動。 此故障偵測與隔離方法之研究目的在於確保自主駕駛車輛能夠安全地運行，本論文將探討電動貨卡複雜的動態行為並將其區分成車輛、輪胎以及動力、轉向、煞車五個部份作分析與建模，車輛控制則包含縱向的巡航控制(cruise control)以及側向的基於目標與控制(Target & Control)模型之道路追隨控制。此研究將使用實驗室近期開發的XiL(X in the Loop)實虛整合模擬驗證系統作模擬與測試，此驗證系統結合ANSYS VRXPERIENCE軟體的虛擬環境、車輛動力學以及dspace輔助開發功能幫助後續的研究工作。

This thesis discusses the vehicle dynamics and subsystems about drive、steering and brake. Developing a fault diagnosis system for actuators to detect sub-system faults when the autonomous vehicle is running. When necessary, the fault source can be identified through the fault diagnosis system. It will correct the control command before the accident occurs. The thesis studies a 3-ton electric truck whose drive system is equipped with a 95kw permanent magnet synchronous motor for rear-wheel drive. Steering system and braking system have been controlled by wire, the system has be driven by servo motor instead of pedal and steering wheel. The purpose of the research of the fault detection and isolation method is to ensure the safe operation of autonomous vehicles. This paper will discuss the complex dynamic behavior of electric trucks and distinguish them into five parts: vehicle, tire, drive, steering and braking for analysis and modeling. The vehicle controls include longitudinal cruise control and lateral lane keeping control based on Target & Control model. This research will be simulated and tested by the XiL (X in the Loop) real-virtual integrated simulation verification system that recently developed by the laboratory. This verification system combines the virtual environment and vehicle dynamics of ANSYS VRXPERIENCE, and dspace software to help research work.