Indoor Vision/INS Integrated Mobile Robot Navigation Using Multimodel-Based Multifrequency Kalman Filter

Abstract

In order to further improve positioning accuracy, this paper proposes an indoor vision/INS integrated mobile robot navigation method using multimodel-based multifrequency Kalman filter. Firstly, to overcome the insufficient accuracy of visual data when a robot turns, a novel multimodel integrated scheme has been investigated for the mobile robots with Mecanum wheels which can make fixed point angled turns. Secondly, a multifrequency Kalman filter has been used to fuse the position information from both the inertial navigation system and the visual navigation system, which overcomes the problem that the filtering period of the integrated navigation system is too long. The proposed multimodel multifrequency Kalman filter gives the root mean square error (RMSE) of 0.0184 m in the direction of east and 0.0977 m in north, respectively. The RMSE of visual navigation system is 0.8925 m in the direction of east and 0.9539 m in north, respectively. Experimental results show that the proposed method is effective.

Predefined-Time Fuzzy Neural Network Control for Omnidirectional Mobile Robot

“…In [34], a sensor fusion method based on the extended Kalman filter is used to achieve path tracking control of a complete mobile robot with four mechanical wheels. Indoor visual navigation of a mobile robot is implemented in [35] using a multi-model, multi-frequency based Kalman filter.…”