The Discrete Event Modeling and Trajectory Planning of Robotic Assembly Tasks

Abstract

A new approach to process modeling, task synthesis, and trajectory planning for robotic assembly is presented. Assembly is modeled as a discrete event dynamic system incorporating both discrete and continuous aspects of the process. The discrete event nature of assembly due to contact state transitions is modeled using Petri nets. The Petri net modeling enables a compact graphical description of causal contact state transitions and provides a coherent mathematical representation of both the discrete and continuous dynamics. In contrast to the traditional contact state network, the Petri net modeling also incorporates causality. Using discrete event modeling, an efficient assembly strategy is found. A discrete event trajectory is determined using dynamic programming to minimize the path length and the uncertainty during assembly. Lastly, an optimal event trajectory is calculated to demonstrate the method. This paper lays the foundation of discrete event modeling for robotic assembly. An new avenue for the analysis and synthesis of significant aspects of the assembly process is opened.