Optimized trajectory planning for the time efficient navigation of mobile robot in constrained environment

Autonomous navigation is a significant segment of mobile robotics, and for reliable autonomous navigation, optimal trajectory planning is the fundamental requirement. In mobile robotics, planning algorithms are implemented to attain optimality in trajectory planning by solving the problems such as path length minimization, smoother trajectories, low computational load, time/ space complexity, etc., that degrade the performance of the path planning technique. This research paper primarily focuses on generating smooth trajectories with the shortest path length by linking the Bidirectional Rapidly exploring Random Tree (B-RRT) with a modified Bezier curve technique termed Smooth-BRRT (S-BRRT). The proposed S-BRRT technique generates smoother trajectories by considering the high number of control points associated with the Bezier curve technique. The selection criteria for control points will be adaptive, which means the number of control points may increase or decrease depending upon the path length, grid cell size, mobile robot dimension, maximum acceleration of mobile robot, etc. The proposed S-BRRT technique is implemented in various simulated environments, and it is experimentally obtained that the path length is reduced by 15.03%, the number of sharp turns is reduced by 100%, and time lag is reduced by 27.01%. The proposed S-BRRT technique is also trialed and tested in various real-world experiments. The result shows a 100% reduction in the collision, the time lag is reduced by 66.23%, and the velocity error is reduced to 57.52%, concerning the results obtained with renowned conventional approaches.