Research on Improved RRT Path Planning Algorithm

Abstract

The traditional RRT (Rapid-exploration Random Tree) path planning algorithm has poor real-time performance and large random adjustment of algorithm parameters, because it requires a large amount of calculations to traverse the entire planning task space. In order to solve the above problems, this paper analyzes the influence on the algorithm performance of the two parameters of step size and selection probability through experiments and statistical analysis of the data. Based on above research results, the principle of algorithm parameter selection and the improved deredundant node algorithm in the process of random tree growth are put forward. Finally, experimental simulation and data collation statistics confirm that the improved RRT path planning algorithm can achieve pruning in the process of random tree growth to make the algorithm quickly converged to the target point, and improves the efficiency of algorithm path planning.