A hybrid noise removal algorithm for MEMS sensors

Abstract The use of microelectromechanical systems (MEMS) in electronics, automotive, consumer and medical sector is increasing rapidly. It is expected that by 2024 the Global MEMS and sensor market will reach the value of 93 billion dollars. MEMS inertial measurement sensors (accelerometer and gyroscopes) are used in low end consumer electronics such as smartphones to high end products such as drones used in military applications. MEMS IMU (Inertial Measurement Unit) are often deployed in dynamic environment such as in machining process for condition monitoring, vehicles for identifying driving behaviors, wearable technology for fall detection and many others. Consequently, Gaussian noise is always present in the output signal of such sensors. Denoising the output signal is critical for feature extraction and data analysis. Recently hybrid approaches for noise removal have gained much research attention. The aim of this paper is to present a hybrid noise removal algorithm for MEMS IMU sensors. In this technique, the signal to noise ratio has been improved and remove glitches by minimizing the hybrid noise from the original output signal of the MEMS IMU sensor.