Enhancing Virtual Assembly Path Planning through a Fuzzy Bayesian Deep Q-Network Algorithm

Abstract

The utilization of virtual reality technology in computer-aided virtual assembly enables the design and planning of assembly processes, enhances production efficiency, and lowers economic costs. Path planning represents a significant avenue of advancement within virtual assembly. Investigating path planning technology within the virtual assembly environment is crucial for developing effective assembly paths in complex settings. This study introduces and evaluates a deep Q-network algorithm based on Fuzzy Bayes within a virtual assembly context. The findings indicate that the Fuzzy Bayes-based deep Q-network algorithm exhibits superior maneuverability and planning efficiency in navigating complex environments with confined spaces.