IQ-RRT*: a path planning algorithm based on informed-RRT* and quick-RRT*

Abstract

Optimal path planning algorithms such as the RRT* and its variants seek to generate the best feasible path from an initial state to a goal state in the least possible time. Prior work on RRT* has focused on improving the convergence rate of the algorithm while keeping its computational complexity unchanged. Informed-RRT* and quick-RRT* are two such variants that, in certain scenarios, converge to the optimal path faster than RRT* does. This work focuses on the novel addition of informed sampling to quick-RRT* to enhance its convergence rate. The resultant algorithm provides initial solutions with costs comparable to quick-RRT* and convergence rates at par with quick-RRT* in the worst case. The authors have concluded that this new algorithm, named IQ-RRT*, outperforms informed-RRT* and quick-RRT* in a multitude of scenarios. IQ-RRT*, unlike quick-RRT*, is a faster alternative to informed-RRT* even in cluttered environments and mazes with long corridors.