Calibration of a micro-electro mechanical system-based accelerometer for vehicle navigation

Abstract

Micro-electro mechanical system-based inertial sensors have broad applications in moving objects including in vehicles for navigation purposes. The low-cost micro-electro mechanical system sensors are normally subject to high dynamic errors such as linear or nonlinear bias, misalignment errors and random noises. In the class of low cost sensors, keeping the accuracy at a reasonable range has always been challenging for engineers. In this paper, a novel method for calibrating low-cost micro-electro mechanical system accelerometers is presented based on soft computing approaches. The method consists of two steps. In the first step, a preliminary model for error sources is presented based on fuzzy subtractive clustering algorithm. This model is then improved using adaptive neuro-fuzzy systems. A Kalman filter is also used to calculate the vehicle velocity and its position based on calibrated measured acceleration. The performance of the presented approach has been validated in the simulated and real experimental driving scenarios. The results show that this method can improve the accuracy of the accelerometer output, measured velocity and position of the vehicle by 79.11%, 97.63% and 99.28%, in the experimental test, respectively. The presented procedure can be used in collision avoidance and emergency brake assist systems.