GA in Continuous Space and Fuzzy Classifier System for Opening a Door with a Manipulator of Mobile Robot: New Benchmark of Evolutionary Intelligent Computing
J. Ohwi, S.V. Ulyanov and Kazuo Yamafuji
The University of Electro-Communications, 1-5-1 Chofugaoka, Chofu Tokyo, 182 Japan
An intelligent mobile robot for service use which is mainly utilized in office buildings has been developed. It can locomote autonomously from one room to others in different floors and buildings using elevators. The robot equips a 5DOF manipulator and must often conduct opening and closing doors operation. For realizing door opening operation of a manipulator in cooperation with locomotive mechanisms, it is necessary to develop efficient and intelligent computing algorithms for next two requirements: (1) determination of initial position of the mobile robot in front of a door-knob like parking a car and (2) path planning of manipulator trajectory for opening a door. This paper provides these algorithms. We propose an algorithm on basis of Fuzzy Classifier System (FCS) for (1). The FCS is a kind of Genetic Algorithms (GA) in which a chromosome expresses each fuzzy rule. And we develop a method of making a trajectory of the flexible manipulator using Genetic Algorithms in Continuous Space (GACS) for (2).