Abstract:

To deal with the problems of slow convergence speed and the irregular searching path of the RRT algorithm, a variable probability search strategy optimization algorithm based on the bidirectional RRT was proposed. This bidirectional RRT algorithm searched different surrounding states of nodes, then extended the target point with different probability strategies. Therefore, this algorithm could rapidly converge to the target in the open environment as well as avoid falling into local minimum. After the initial path search was completed, the irregular route was optimized according to the location of the path node, and the number of turning times during the car driving and the length of the total path were effectively reduced.The multi scene repeated test was carried out in the simulation, and the simulation results showed that this algorithm had a significant improvement in the search speed and path length.

Key words: intelligent vehicle, path planning, bidirectional RRT, probability search, convergence speed, irregular searching path