Implementation of a Multi-Layered Fuzzy Controller on an FPGA
Abstract (Summary)The uses of Autonomous Mobile robots range from the teleoperated Sojourner on the Mars Pathfinder mission to cleaning robots in the Paris Metro. They have overtaken tasks which are too difficult or dangerous to perform or require the repetition of the same procedure over and over again and can become dull for a human to perform. The objective of this thesis is to develop and implement a multilayered fuzzy controller on an FPGA for the navigational control of an autonomous mobile robot (AMR). The navigational algorithm assists the robot in navigating through complex environments of passages and hallways. The FPGA implementation of the navigational algorithm employs a layered fuzzy control strategy where a control model is developed using three elementary layers of control: Centering Layer, Wall Hugging Layer, and Obstacle Avoidance Layer. The centering layer is used to control the robot motion through a passage with both walls within the range of robot side sensors. In such a scenario, this layer directs the robot to the center of the passage. If the side sensors for the robot detect wall only on one side, wall hugging layer is activated and the robot moves parallel to the side wall. These two layers are developed using a proportional-plus-derivative fuzzy controller and share the fuzzy sets and the rule base. The obstacle avoidance layer prevents the collision of the robot with objects and is activated if the front sensor detects an obstacle within a preset distance.
School:Wright State University
School Location:USA - Ohio
Source Type:Master's Thesis
Date of Publication:01/01/2006