A Sliding-Mode Control Algorithm of Self-Learning and Obstacle Avoidance for Mobile Robots

It is generally very difficult to make effective obstacle avoidance for mobile robots, especially in uncertain environments. This article models the obstacle avoidance problem as a nonlinear sliding mode, using the self-learning method of non-linear system. First, a sliding mode algorithm is proposed for state dependent layers of the mobile robots, in which two kinds of boundary layers are included in, namely a sector-shaped layer and a constant layer. Second, a multi-input algorithm based on sliding-mode for self-learning of mobile robots is discussed. Some control rules are built for the self-learning and obstacle avoidance of mobile robots, and the solving steps are also presented. Last, an experiment is designed to verify the proposed model and calculate the slidingmode control for mobile robots. Some interesting conclusions and future work are indicated at the end of the paper.