Rover Trajectory Planning via Simulation Using Incremented Particle Swarm Optimization

This paper presents an off-line optimal trajectory planning for differential-drive rover through simulation of the dynamic model. The paper starts with the model dynamics of an actual rover built in our space science and technology lab (SSTLab) and controlled by simple PD controllers. Next, the proposed optimization technique used is presented which is called Incremented Particle-Swarm Optimization (IPSO) where the number of variable increases incrementally if the goal is not satisfied to minimize the time and CPU usage running the cost function. Different trajectories and cost functions were tested with obstacles and without it. The results show that the trajectory can be optimized efficiently using IPSO and a simple cost function based on total time and distance to final destination.