General Type-2 Fuzzy Reasoning for Path-Planning of a Mobile Robot in a Dynamic Environment under Sensory Uncertainty

Abstract

Memberships assigned by an expert to different values of a linguistic variable in a classical (Type-1) fuzzy set are crisp. A Type-2 Fuzzy membership function is a union of several Type-1 membership functions obtained from different experts. Type-2 membership at a given value of a linguistic variable is thus fuzzy. Traditional path-planning problem of a mobile robot in a static environment includes Type-1 (sensory) uncertainty due to the angular spacing betweeneach pairs of consecutive ultrasonic sensors mounted around the body of the robot. Type-2 uncertainty appears in path-planning problems due to the presence of moving obstacles in the robot’s workspace. Membership functions used for handling type-1 sensory uncertainty are inadequate to handle type-2 uncertainty in a dynamic environment. This paper introduces a novel algorithm for automated reasoning with the help of a vertical structured General Type-2 version of fuzzy sets, and extends it for trajectory-planning of a mobile robot in presence of dynamic obstacles. A set of 7 benchmark workplaces is used to test the performance of the proposed algorithm, taking specular reflection and multipath effect of sonar transducer into consideration. It is noted that the proposed algorithm outperforms existing techniques with respect to path length and computational overhead.