Three-Dimensional Angles Measurement using IME based on MEMS Technology by Adaptive Kalman Filter

Three-dimensional angular measurements are used in a large number of applications, such as           positioning systems, robotics, motion analysis, control analysis, and so on. The sensors used for this       purpose are the accelerometer and the gyroscope. The angle measurement in the accelerometer sensor is based on the vector component of the gravitational acceleration on the sensor axes and the trigonometric relations. For this reason, only a quasi-static limit can be obtained from the almost precise angle. In      dynamic mode, angle measurement using this sensor is accompanied by large errors, due to the effect of other accelerations, in addition to gravity acceleration. The angle can also be obtained by integrating the gyroscope output, but in practice due to the problem of gyroscope drift, the measurement angle is far from the actual value, with increasing error over time. In this paper, we have been able to cover the weaknesses of each of these sensors using a combination of accelerometer and gyroscope features by an adaptive     Kalman filter. The proposed adaptive filter is able to adjust itself to the static and dynamic conditions, so that it can combine the output information of the sensors properly. To evaluate the proposed structure,    dynamic and static experiments were conducted on an IMU based on MEMS technology. The results show that compared to ordinary Kalman filter, the RMS error of angle measurement in the proposed scheme has an improvement of 34% in the static mode and 34.3% and 29.8% in dynamic modes of pitch and roll,     respectively.