High precision odometer estimation algorithm for household intelligent nursing robot

Abstract

In order to solve the problem that the imbalance between wheel tire of home nursing robot and the ground force and the motor output torque is easy to the sliding of the robot and the low accuracy of odometer estimation,the traction coefficient was introduced to describe the sliding condition of the robot,and the kinematic model of robot containing traction coefficient was deduced.Two sensors of incremental photoelectric encoder and inertial measurement unit were adopted as inputs to apply the kinematic model containing traction coefficient in the odometer estimation algorithm based on extended Kalman Filter algorithm.Finally,a robot experimental system was formulated to complete the algorithm verification.The results indicate that when the robot moves at the speed of 0.1,0.2,0.4 m/s on home ceramic ground,the estimation error of the proposed robot odometer is reduced by about 40% compared with the traditional odometer estimation algorithm.Applying the kinematic model with traction coefficient to the robot range meter estimation algorithm can effectively reduce the odometer estimation error of the nursing robot,thus providing a certain reference for the improvement of the independent navigational accuracy of the home nursing robot on indoor grounds.

Keywords

• robot control; intelligent nursing robot; odometer estimation; multisensory fusion; traction coefficient