Multi-sensor attitude algorithm design for a low-cost strap-down system based on the direction cosine matrix

ABSTRACT:   In order to satisfy the urgent needs of low-cost, high-performance measurement system for the attitude angle of the carrier, this paper presents a multi-sensor data fusion algorithm based on Direction Cosine Matrix (DCM). Aiming at the problem of inaccurate measurement of the attitude incurred by low-cost MEMS gyroscope drifting, the study constructs a multi-axis complementary filter, proposes reference vectors to detect gyroscope bias error in each axis and uses PI controller as feedback to compensate the drift error of gyros. Static experiment on a three-axis turntable shows that the static accuracy of the attitude angle is better than 0.2°, and dynamic vehicle experiment shows that the dynamic accuracy is better than 4°, meeting the application requirements of low-cost attitude determination system.