An optimal sampling-based path planning under uncertainty based on linear quadratic regulator

The navigation of mobile robots is affected by noises from perceptual systems and positioning systems, as well as track deviations from control systems. Therefore, it is significant to ensure the stability of path planning in an uncertain environment. In this paper, an optimal samplingbased path planning method based on linear quadratic regulator is proposed. This method uses heuristic rapid-exploring random tree to generate paths which are executable and asymptotic optimal. Then, the priori distributions of the generated paths are calculated and evaluated through an uncertainty factor. The effectiveness of the proposed method is verified by simulation experiments in a variety of uncertain environments.