Intelligent route planning for fast autonomous vehicles operating in a large natural terrain

This paper presents a 2D route planner to support operations for an autonomous vehicle conducting missions at high speeds in a large natural terrain. A configurable search engine is used to search a pre-prepared digitized map of the natural terrain and to generate a suboptimal route in terms of distance, safety, and maneuvering. The suboptimal route is more optimized and smoothed by a route filter that produces the optimal route under some conditions. A high-level intelligent supervisory module configures the search engine and assigns weights to the three route cost components to satisfy the mission objectives and the time constraints of the mission. To take advantage of previous route planning experience, an intelligent learning support engine is used to collect route cost data and to build a knowledge base that is used to provide better heuristics to the search engine. The presented approaches and results contribute towards a truly autonomous vehicle capable of conducting complex missions efficiently and in real time. (C) 2002 Elsevier Science B.V. All rights reserved.