Path planning of target search for mobile robot with expected time

In the target search problem of mobile robots under uncertain environment, the probability that a target is found at an observation point is often set to an ideal uniform distribution, the path optimization index is usually the shortest distance, and the shortest distance path is not equivalent to the optimal expected time path. For this issue, a probabilistic multi-target search algorithm with expected time as index is proposed in this paper. Aiming at the phenomenon different access order of nodes will lead to different expected-time, a hierarchical path optimization strategy is applied in path planning. First, a new probability estimation model for non-uniform target distribution is constructed; and then an improved circle modification algorithm is employed to generate the optimal expected observation point sequence at top level planning; Finally, the goal biasing collision rapid-exploration random tree (GBC-RRT) algorithm is used in observation points at lower level planning to realize a feasible path under feature maps. Simulation experiment results show that the method proposed significantly reduces the expected-time for the mobile robot to search the target, and can obtain the optimal expected-time path planning under uncertain and non-uniformly distributed working environments. © 2020, Editorial Department of Control Theory & Applications South China University of Technology. All right reserved.