A Real-Time and Accurate Coordinate Positioning Model Based on Gyroscopes and Omni-directional Driven Wheels

Abstract

A novel simple efficient modular coordinate positioning model, based on the study of position estimation for mobile robots, is raised. It uses orthogonal omnidirectional driven wheels to record displacement instead of accelerometers in the inertial positioning, solving the problem that displacement error calculated from the quadratic integral of the accelerometer is too big. It combines with the yaw angle gained from the gyroscope to calculate accurate coordinates. Considering the problem that the model depends on the ground flatness a lot, two optimization methods are brought up, namely Rotation Redundancy Elimination and Uneven Ground Compensation. In this way, the model can has more accuracy. Experimental results show the model is simple and real-time, and does have high positioning accuracy and strong applicability.