Motion planning algorithm for tractor-trailer mobile robot in unknown environment

Abstract

A path planning method for the tractor-trailer mobile robot in unknown environment based on rapid-exploring random trees is proposed. The space model of the environment is firstly constructed with sonar sensors. Mechanism of evaporation and enhancement are designed to reduce the dimension of the search space. The feasible path is generated with Dual-RRT algorithm, which is satisfied with the nonholonomic and mechanical constraints. Disbranch operation on the RRT is executed to avoid the repeated searches and increase the search speed. Results of simulation and pro-type experiment show that the tractor-traile mobile robot can quickly plan a feasible path under complex environment, which not only avoid the any collision with the obstacles but also satisfies the constraints of the system.