Cubature Kalman filter with risk sensitive cost function

A novel method to optimize risk sensitive cost function for nonlinear system based on cubature quadrature rule has been proposed in this paper. The proposed filter has been named as risk sensitive cubature Kalman filter (RSCKF). Also a simple and easy to follow derivation of cubature quadrature rule for multi dimensional integral has been provided. Although the computational load is comparable with extended risk sensitive filter (ERSF), the proposed filter is able to overcome the inherent disadvantages associated with it. The theory and formulation of proposed RSCKF have been presented in this paper. Due to more accuracy, enhanced robustness and computational efficiency compare to ERSF, the proposed robust estimator may find place for on-board real life applications.