Design of the self-optimal tuning fuzzy inference system for transformer network-based traveling salesman problem in tidying up objects

This research focuses on optimizing the path for service robots to tidying up objects in real-world environments. The study involves converting the physical field containing objects requiring the robot’s attention into virtual data. Multiple layers of environmental information are derived from this virtual data to analyze and optimize the path for robot-assisted tidying. The Mountain Gazelle Optimizer is employed to enhance the fuzzy inference system’s rule base, resulting in improved performance. Comparative analyses with alternative metaheuristic algorithms confirm the effectiveness of the Mountain Gazelle Optimizer in enhancing the fuzzy inference system. Additionally, a Transformer Network is utilized to identify the optimal path for tidying objects. This study significantly enhances the tidying capabilities and efficiency of service robots, contributing to real-world applications.