An Enhanced Genetic Algorithm for Solving Trajectory Planning of Autonomous Robots

Different tasks inside the environments call for different types of robots with various traits and designs. The research community appears to favour a number of robot classes in particular area for tackling complex challenges in demanding settings. Important cases of these deployments are examined, and it is determined that high-level autonomy is a crucial problem that has to be solved. It is necessary to drive autonomous mobile robots across the environment to map it, find themselves, and chart out routes between locations. Two crucial capabilities of an autonomous overtaking system are trajectory planning and trajectory tracking, and many approaches have been put forth in the literature for each of these functionalities. However, the majority of the offered approaches are only useful for low-speed overtaking due to ambiguities in environment perception when employing the current generation of sensors. This paper shows an upgraded version of Genetic Algorithm to find the shortest route between the source and destination. In this paper, an upgraded GA is proposed in static environments to solve path planning issues. Numerous studies have offered innovative methods that generate an ideal route using GA. Most of the strategies for generating infeasible pathways ignore the variable length chromosomes. The algorithm converges more quickly because it prevents premature convergence and provides viable pathways with higher fitness values than its parents. The suggested approach is implemented and contrasted in a variety of contexts to demonstrate its validity. The simulation findings demonstrate that, in comparison to previous approaches, utilizing GA with the enhanced operators and the fitness function aids in the discovery of optimal solutions. It is clear from the best optimal solution provided that the proposed UGA outperforms GA. In UGA the execution time is less (15.24s) than the GA that is (19.35s). The proposed model requires less iterations (3127) than the current method (6114), which reduces the total number of iterations. This in turn cuts down on the overall execution time.