A Hybrid PSO-GWO Algorithm for Robot Path Planning in Uncertain Environments

Abstract

Path planning is one of the most critical tasks in mobile robotics. This paper proposes a hybridization of two algorithms for optimizing the trajectory planning a mobile robot and ensuring its navigation in a partially known environment containing static obstacles. The Grey Wolf Optimization (GWO) algorithm, which has some challenges such as slow convergence and low accuracy, is combined with the Particle Swarm Optimization (PSO) algorithm for its high convergence speed and accuracy. The hybrid PSO-GWO algorithm is proposed so that the optimal trajectory search procedure balances global and local search capabilities, taking advantage of the capacities of both GWO and PSO approaches. The validity of PSO-GWO performance has been tested for various scenarios using the mobile robot Khepera-iv and the simulation software of MATLAB and V-REP. Then, it has revealed that this method can offer a better search for an optimal trajectory when compared to other algorithms such as GWO, PSO, and Fuzzy-Logic. For implementing the PSO-GWO algorithm on Khepera-iv, a simple Extended Kalman Filter (EKF) is developed to locate the mobile robot, in its realistic environment, by fusing data from different sensors.