To address the issues of solid randomness in sampling points

poor path directionality

low path smoothness

and long paths in high-precision path planning of rapid-exploration random tree (RRT) algorithm

a fused artificial fish swarm algorithm (RRT-ASFA) is proposed to optimize the paths generated by RRT. Firstly

a target biasing strategy is proposed for RRT to reduce the randomness of sampling points and optimize the target direction; and step size adaption and search area restriction are proposed to optimize the path planning time. For the artificial fish swarming algorithm (ASFA)

the adaptive step size and adaptive field-of-view range are then proposed to make the artificial fish swarm converge faster

and the turning points of the paths planned by RRT are optimized to make the paths shorter. Finally

the paths are smoothed by the Hermite spline function. Through simulation experiments

it is found that the combined algorithm demonstrated shorter path length and fewer path nodes compared to traditional RRT

target-biased RRT

and RRT* algorithms

proving the combined algorithm's feasibility.