Prediction-based WSN-mediated dynamic obstacle avoidance for mobile robot

Utilizing the advantages of wireless sensor network (WSN), this paper puts forward a novel dynamic obstacle avoidance algorithm used in unknown complex environment, which is a fundamental problem and important research area of mobile robots. In view of moving velocity and direction of both the obstacles and robots, a mathematic model is built based on the exposure model, exposure direction and critical speeds of sensors. A position prediction algorithm is introduced in order to improve the accuracy and robustness of the system for tracking moving obstacles using a Kalman filter, following the principle of least standard deviation of Kalman predication curve from practical curve. A practical implementation with real WSN and real mobile robots has been carried out to validate the enhanced efficiency, stability and accuracy of the proposed algorithm for dynamic obstacle avoidance in real time.