Implementation of A low-cost multi-IMU hardware by using a homogenous multi-sensor fusion

In this paper a homogenous multi-sensor fusion method used to estimate true angular rate with combination of four low cost MEMS Inertial Measurement Unit (IMU) in order to reduce noise effects. A discrete-time Kalman filter designed to combine output of four low accuracy sensors with price around 10$ to give more exact rate value. A hardware implemented to test the method and results show the improvement in variance of estimated angular rate more than 5 times and ARW error reduction about 2 times respect to the single gyro. A comparison between Kalman filtering and a simple averaging method was performed that shows the improvement of accuracy can be more than the averaging method.