Noise Removal in The IMU Sensor Using Exponential Moving Average with Parameter Selection in Remotely Operated Vehicle (ROV)

Abstract

The presence of vibration noise in Remotely Operated Vehicle (ROV) is inevitable. This vibration causes noise in the Inertial Measurement Unit (IMU) output. This noise can be filtered out using a simple Exponential Moving Average (EMA) filter. However, the performance of the filter depends on the value of $a$. In this research, an approach to select $a$ is proposed. Based on the ROV implementation that maintains its position under the water using $a$ = 0.031, the root mean square error (RMSE) for 4 trials is 0.19, 0.22, 0.29, and 0.75. Time consumptions for a single filtering process are also small, which is 0.047 seconds. It can be concluded that the proposed approach can be used for ROV to pre-process the output of the IMU sensor prior to proceeding to further process.