A New Technique for Optimization of Linear Displacement Measurement based on MEMS Accelerometer

Linear displacement has been widely applied in the technology and applications for Industry 4.0. The Micro-electromechanical systems (MEMS) accelerometer is a potential candidate in linear displacement measurement by means of acceleration acquisition. However, the error still presents on the displacement measurement at the starting point or each time the sensor ends its motion and returns to a static state due to the noise and imperfection of mechanical structure. The paper presents a new algorithm, 'No Displacement Zero Translational Accumulation' (NDZTA), to solve the cumulative error of linear measurement when the sensor stays at stationary points. The proposed filter removes noise by pulling down the current velocity and acceleration to 0 immediately when the sensor finishes its motion. Moreover, a simple but practical way to calculate the displacement via the 'Area Integration Method' was visualized and analyzed in detail. The experimental result shows a good accuracy as well as stability of measured result with the maximum error only about 0.2 cm in the range of 200 cm.