Designing Type-2 Fuzzy Behaviors of Autonomous Robot Using Interval Type-2 Fuzzy Logic System

抄録

The type-2 fuzzy set being an extension of original fuzzy set was introduced by L.Zadeh in 1975 and J.Mendel has developed Type-2 Fuzzy Logic System (FLS). Type-2 FLSs are better than type-1 FLSs on management of uncertainty. This paper deals with an approach in designing fuzzy behaviors of an autonomous robot navigation system using interval type-2 FLS. The navigation system uses sonar sensors to detect obstacles and model the environment. When detecting a obstacle, the fuzzy spatial relationship between robot and the obstacle is computed to be an input of interval type-2 FLS. Authors also proposed the hierarchy of behaviors to fuse command from fuzzy behavior and defuzzify to get crisp angle of deviation. The simulated results on graphic environment and some statistics are shown to demonstrate our approach is implemented correctly.