MEMS based IMU for tilting measurement: Comparison of complementary and kalman filter based data fusion

This research investigates real time tilting measurement using Micro-Electro-Mechanical-system (MEMS) based inertial measurement unit (IMU). Accelerometers suffer from errors caused by external accelerations that sums to gravity and make accelerometers based tilting sensing unreliable and inaccurate. Gyroscopes can offset such drawbacks but have data drifting problems. This paper presents a study on complementary and Kalman filter for tilting measurement using MEMS based IMU. The complementary filter algorithm uses low-pass filter and high-pass filter to deal with the data from accelerometer and gyroscope while Kalman filter takes the tilting angle and gyroscope bias as system states, combining the angle derived from the accelerometer to estimate the tilting angle. The study carried out both static and dynamic experiments. The results showed that both Complementary and Kalman filter were less sensitive to variations and almost no signal coupling phenomenon and able to obtain smooth and accurate results.