Two-step optimal filter design for the low-cost attitude and heading reference systems

This study presents a novel sensing methodology with two optimal condition-based fusion algorithms for attitude estimation, using low-cost micro-machined gyroscopes, accelerometers and magnetometers. The proposed methodology named two-step optimal filter is composed of an optimal filter and fast determination algorithm. The filter is designed as sensor-based Kalman filter, which is augmented by a fuzzy rule to adjust the parameters on line to yield optimal measurements of accelerometers and magnetometers. Then, the fast second estimator of the optimal quaternion algorithm is described to determine the orientations. Meanwhile, adaptation architecture is implemented to yield robust performance, even when the vehicle is subject to strong accelerations or ferromagnetic disturbed. The new construction of attitude estimation algorithm is easy to be implemented, the precise, robustness and efficient are compared with the common methodology. Experimental results are provided for a remotely operational vehicle test and the performance of the proposed filter is evaluated against the output from a conventional filter.