eduzhai > Applied Sciences > Engineering >

Computationally Efficient Attitude Estimation with Extended $\mathcal{H}_2$ Filtering

  • Save

... pages left unread,continue reading

Document pages: 22 pages

Abstract: Accurate state estimation using low-cost MEMS (Micro Electro- MechanicalSystems) sensors present on Commercial-off-the-shelf (COTS) drones is achallenging problem. Most UAV systems use a combination of a gyroscope, anaccelerometer, and a magnetometer to obtain measurements and estimate attitude.Under this paradigm of sensor fusion, the Extended Kalman Filter (EKF) is themost popular algorithm for attitude estimation in UAVs. In this work, wepropose a novel estimation technique called extended H2 filter that canovercome the limitations of the EKF, specifically with respect to computationalspeed, memory usage, and root mean squared error. We formulate ourattitude-estimation algorithm using unit quaternions. The H2 optimal filtergain is designed offline about a nominal operating point by solving a convexoptimization problem, and the filter dynamics is implemented using thenonlinear system dynamics. This implementation of this H2 optimal estimator isreferred as the extended H2 estimator. The proposed technique is tested on fourcases corresponding to long time-scale motion, fast time-scale motion,transition from hover to forward flight for VTOL aircrafts, and an entireflight cycle (from take-off to landing). Its results are compared against thatof the EKF in terms of the aforementioned performance metrics.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×