Hardware/Software Codesign for a Fuzzy Autonomous Road-Following System

摘要

In this research, a fuzzy logic controller is designed for vision-based autonomous road-following. Because of its high-speed response, portability, and flexibility, a field programmable gate array is applied to implement this control system. Furthermore, a novel hardware/software partitioning method using the genetic algorithm is developed. This method is capable of finding the tradeoff among several evaluation factors under conditions of hard constraints. A small-scaled intelligent vehicle, which is capable of autonomous road-following is built and the proposed controller is tested in the real-world environment. Experiments of hardware implementation and codesign implementation are presented and compared.