The automated lane-keeping design for an intelligent vehicle

Abstract

In this paper, a vision-based lane-keeping automated steering system is proposed and is successfully verified in our vehicle platform, TAIWAN iTS-1. The proposed steering system can achieve the accurate detection of the complicated road environment information; and more, the closed-loop automated lane-keeping steering system with virtual look-ahead is stable under varying speed operation. Furthermore, to achieve more manlike driving behavior such as smooth tuning, a fuzzy gain schedule technology is proposed to concern with lateral offset and instant-speed of the vehicle, and hence, to compensate the feedback controller for adapting to the steering wheel command appropriately. The proposed steering system is demonstrated via TAIWAN iTS-1 on the standard testing road in automotive research and testing center (ARTC) and highway road.