Using ABC and RRT algorithms to improve mobile robot path planning with danger degree

Abstract

For mobile robots, being able to find a suitable route through an environment filled with varied obstacles, and to ensure that they can successfully reach their target point from the starting point in the most efficient manner is very important, and a necessary research topic. This article proposes a combination of Artificial Bee Colony Algorithm (ABC) and Rapidly-Exploring Random Tree (RRT) to produce a novel algorithm to meet these navigation requirements. This algorithm is then compared with the traditional ABC for path planning. Unlike previous algorithms, this study uses the RRT algorithm to find several extend points, choose the best extend point to move the bees. Because the Artificial Bee Colony algorithm is simply structured, easy to operate and quickly converges, it is able to address the problems of slow convergence and easy entrapment in local optimal solutions encountered in previous path planning algorithms. Although RRT has excellent characteristics in terms of the search area which is unknown, it is unstable for each planning. Thus, this thesis combines the characteristics of the artificial bee colony algorithm with the RRT algorithms, and considers the problem of robot path simulation with obstacles.