Guaranteed Cost Robust Modified Covariance Intersection Fusion Kalman Filter for Multi-Sensor System with Uncertain Noise Variances and Random Missing Measurements

摘要

This paper deals with the design of guaranteed cost robust modified Covariance Intersection (CI) fusion Kalman filter for time-invariant multi-sensor systems with uncertain noise variances and random missing measurements. The systems are converted into that with only uncertain noise variances by introducing the fictitious noise. According to the mini-max robust estimation principle and Lyapunov equation approach, the two classes of guaranteed cost robust modified CI fusion Kalman filters are proposed. They are converted into corresponding nonlinear and linear optimization problems by the parameterized representation of uncertain noise variance perturbations and can be solved by the Lagrange multiplier method and linear program (LP) method, respectively. Applying the Lyapunov equation approach, the guaranteed cost robustness is proved. A numerical example is given to verify the correctness and effectiveness of the proposed results.