In order to solve the problems of RRT*-Connect algorithm in complex obstacle environment

such as large randomness

low search efficiency and redundant path nodes

a path planning algorithm of mobile robot based on dynamic constrained sampling RRT*-Connect algorithm is proposed. Firstly

based on RRT*-Connect algorithm

dynamic target node probability bias is used to sample the target node of the opposite search tree

so that the target of double tree extension is the node of the opposite search tree and it reduces the randomness of the double tree connection. Combined with the dynamic region sampling method of target node

the multi-stage sampling region is constructed with the target node and the node nearest to the target node as the core

and the expansion direction of the search tree is constrained to realize dynamic constrained sampling

and the local oscillation is avoided by setting a threshold. On this basis

the repulsive force idea of artificial potential field method is introduced to design the adaptive dynamic step size based on the global and local environmental complexity coefficient

which greatly improves the search efficiency. Additionally

Cantmull-Rom spline curve was used for path smoothing. The simulation results show that the algorithm has faster convergence speed

higher efficiency

more than 30% node utilization rate

significantly reduced useless nodes

less searching branches

and can adapt to various complex obstacle environments. Compared with RRT*-Connect algorithm

the running time of the proposed algorithm is reduced by more than 23%

and the path length is reduced by more than 7%.