Enhancing Path Planning of Assistive Robots in Complex Environments Using Geno-Fuzzy Algorithm

Abstract

The path-planning algorithm is the central part of most v. The algorithm should consider fixed obstacles, furniture and building style, dynamic obstacles, humans, and pets. assistive robots encounter a challenging and complex environment with various obstacles during daily work. In addition, to maximize the service per hour, the robot has to select the optimum path. These challenges motivate the work toward an efficient path-planning algorithm that can handle complex environments. The proposed algorithm employs a designed genetic algorithm to look for the best path that maximizes the service area per hour. This genetic algorithm is then combined with a dynamic obstacle detection fuzzy system. This system relies on fuzzy membership zones. The algorithm decides whether the obstacle is dynamic or static according to speed, direction, and size. The Geno-fuzzy path planning algorithm is implemented in an assistive robot and tested in an actual environment. The algorithm implementation in a simulated environment of 100 BED hospitals in Iraq reveals a high-performance result. The test on a large scale without obstacles shows the ability of the algorithm to deal with more than 300 service points successfully. The local experiment on Webots proved the algorithm's performance to overcome dynamic obstacles and achieve safe traveling.