自駕車超車控制及避障策略開發

Development of Overtaking and Obstacle Avoidance Control Strategy for Autonomous Vehicles

李則霖 ； 盧玟翰 ； 彭毓瑩 ； 鄭榮和

《機械工業雜誌》 409期 (2017/04) Pp. 70-83

摘要

由於對車輛安全的重視程度持續增加與先進駕駛輔助系統(advanced driving assistance system, ADAS)技術漸趨成熟，自動駕駛車在近年來成為熱門研究領域，各家車廠相繼投入，試圖以各自方式解決交通安全議題。為了要能應用在一般道路上，自動駕駛車除了要能從指定的起點(依指定路線)開到目標終點外，更要解決過程中可能會遇到的一系列交通問題(舉凡交通號誌辨識、行人/障礙物辨識、超車控制與避障等)。本文針對超車控制與避障議題，以工研院機械所智慧車輛組開發之智慧四輪電動車(CPEV)為對象，利用MATLAB/Simulink建立控制策略模型，搭配以CarSim軟體建立之車輛動態模型進行模型迴路模擬(MIL)驗證與策略參數調整，開發適當、安全的超車避障策略，避免超車過程中發生危險。

關鍵字

超車控制 ； 避障 ； 自駕車

並列摘要

Since vehicle safety is increasingly emphasized and Advanced Driving Assistance Systems (ADAS) technology is getting more and more matured, self-driving of vehicles has become a hot research field in recent years. Many automobile manufacturers take part in this trend and try to solve road safety issues on their own ways. To bring the autonomous vehicles to the real road application, the vehicles needs not only the ability to drive from the designated starting point to the target end (according to the designated route) but also the ability to solve a series of traffic problems that it may encounter on route, such as traffic light recognition, pedestrian and obstacle identification, overtaking and obstacle avoidance. In this article, a control strategy model and a vehicle dynamic model for the intelligent electric vehicle (called CPEV) developed by ITRI are established on MATLAB / Simulink and CarSim. Through Model-In-the-Loop (MIL) simulation and parameter tuning, an overtaking and obstacle avoidance strategy which can help vehicles safely avoid danger in the process of overtaking has been developed.

並列關鍵字

Overtaking Control ； Obstacle Avoidance ； Autonomous Vehicles