Using Fuzzy Logic in Automated Vehicle Control

Abstract

The automatic-driving field has received much attention in recent years, as exemplified by the Darpa's Grand Challenge. Two Spanish research groups have furthered such work by automating two mass-produced vehicles. As input, their system uses a centimetric global positioning system, wireless LAN support, and artificial vision. To control the vehicle, they use fuzzy logic techniques that contend with both complex mathematical models and inaccurate linearization. Fuzzy logic also lets them incorporate human procedural knowledge into their control algorithms. Here, the researchers describe the algorithms for steering and speed control, which together make up the trajectory control. They also describe algorithms for overtaking, adaptive cruise control with stop-and-go functionality, and vision-based vehicle detection, and discuss results from experiments in real-world conditions.