Path-planning for mobile robot based on data fusion and artificial potential
Min Han Tan · Journal of Shandong University · 2005
Robot path planning and motion control in dynamic uncertain environment are studied. Firstly, in time domain, ultrasonic data predictive model is proposed, and the predictive model parameters are updated using the Recursive Least Squares method. By fusing the predictive and the actual values of the ultrasonic sensor, the reliability of the ultrasonic data is improved remarkably. Considering the consistency of the environment, adjacent ultrasonic data are fused to improve the practicality of the sensor data. Taking advantage of the rolling optimization concept adopted in predictive control, the problems of robot navigation and autonomous obstacle avoidance are resolved by using artificial potential field due to its simplicity. The effectiveness of the proposed method is proved by experiments on the mobile robot CASIA-1.