Fuzzy predictive control of wheeled mobile robot based on multi-sensors

Abstract

During the wheeled mobile robot (WMR) navigation, the position estimation obtained by one sensor is unrealistic and useless. The composite navigation method based on multi-sensor is proposed, which provides redundancy and assures reliability and precision of the observed features. Considering the non-holonomic constraint between the ground and WMR, a fuzzy predictive control is presented where the weight matrix of the pose error is chosen by the real-time fuzzy algorithm. Here, the fuzzy predictive controller based on composite navigation is used on XUAT.AGV, which is a kind of WMR. It has been proved that the controller can improve the control reliability and precision of WMR through theory analysis and experiment.