Low Cost System: GPS/MEMS for Positioning

This paper presents results of tests on a system which uses Micro-Electro-Mechanical-Systems (MEMS) and global positioning system (GPS) in place of more expensive Inertial Navigation Systems (INS). This is not an entirely new approach to sensor integration for positioning, as using a number of MEMS to replace INS has been studied since these components were put into the market. In our case we apply an algorithm for integrating data from GPS and MEMS in real time placed inside a van for mobile mapping. The Extended Kalman Filter (EKF) was used for this purpose. The method assumes movement on a plane, and analyses how EKF can reduce error thus decreasing the accuracy difference between INS and GPS/MEMS. It is taken into account that MEMS are not calibrated, therefore systematic errors which are not compensated by the EKF will occur. It is necessary to modify the EKF model putting new variables that represent error variables to account for bias and scale errors in the MEMS