Modeling of sensor-based robotic task plans using fuzzy Petri nets

This paper proposes a fuzzy Petri net (FPN) model of sensor-based execution of robotic task plans. A fuzzy Petri net consists of three types of fuzzy variables: local fuzzy variables, fuzzy marking variables, and global fuzzy variables, which are used to model object attributes, system states, and degrees of completion of global tasks. An executable fuzzy Petri net, a subnet of the FPN for sequence planning, is obtained using a feasible, complete, and correctly ordered sequence. An example which incorporates modeling and error detection and recovery using these concepts is given