Dynamic Attitude Estimation Using MEMS and Optical Flow Sensors

Abstract

Micro-electro-mechanical system (MEMS) sensors have been widely used in attitude and heading reference systems (AHRS), including MEMS accelerometer and MEMS gyroscope. The accelerometer can measure gravity vector and help to determine pitch and roll angles, but its measurement is actually the sum of gravity and motion acceleration. As a result, the accuracy of pitch and roll angles will be affected under dynamic conditions. In this paper, optical flow sensor is used to measure the linear motion, and thus the motion acceleration can be estimated from optical flow data. Experimental results show that the proposed method can effectively eliminate the impact of motion acceleration and enhance the dynamic accuracy of AHRS.