Range‐parameterised orthogonal simplex cubature Kalman filter for bearings‐only measurements.

In this study, a non‐linear filter named the range‐parameterised orthogonal simplex cubature Kalman filter (RPOSCKF) is proposed to further improve the accuracy of bearings‐only tracking. The filter combines the simplex spherical radial numerical rule with an orthogonal method, and thus the standard cubature points are transformed into the orthogonal simplex cubature points (OSCPs). It is proved that the filter using the OSCPs can tune the high order terms of the highly non‐linear measurement function and therefore alleviate problems due to the non‐local sampling effects. On the other hand, the fuzzy initial estimation problem is handled by an improved range‐parameterised (IRP) strategy. The IRP strategy divides the filter into some weighted orthogonal simplex CKFs each with different initial estimate, where the initial weights are based on the length of the estimate interval. Also, the sub‐filters with small updated weights will be removed and the computational complexity can be reduced. Simulations show that the proposed algorithm provides improved performance over the conventional algorithms. [ABSTRACT FROM AUTHOR]