"Fuzzy Logic: Autonomus Robot Navigation with FuzzBot"
Brian Conaway 2006
Abstract
This project examines fuzzy logic and its practical use in mobile robotics. It applies fuzzy logic control to a sensor based robot, FuzzBot. The fuzzy controller constucted in this project provides a mechanism for obstacle avoidance. The fuzzy logic controller raises the issue of whether or not FuzzBot can effectively adapt itself to autonomous environments.
To solve this issue, the project has designed a fuzzy logic controller and implemented it on an OOPic OOBoard to achieve obstacle advoidance behavior. FuzzBot receives distance readings from an infrared sensor, fuzzifies them, evaluates them, and slects appropriate rules for output. Having selected the rule, it defuzzifies the output intp crisp values to select an output action for FuzzBot.
Scenarios are created to test the correctness of FuzzBot's fuzzy logic controller. FuzzBot is evaluated based on how efficiently and successfully it completes each senario. The senarios provide a representative sample of the possible problems FuzzBot will face in any environment. Results from the analysis conclude that a fuzzy logic controller allow FuzzBot to adapt to autonomous environments and successfully avoid obstacles.