A fuzzy adaptive fusion algorithm for radar/infrared dual mode guidance

An adaptive decentralized fusion algorithm based on H_∞ filter and fuzzy adaptive fusion is proposed for radar/infrared dual, mode guidance system, which combines the merits of H_∞ filter and fuzzy logic. To overcome the defect of the dependence on priori assumption of noise covariance matrices of Kalman filter, H_∞ filter is introduced, which needs no priori knowledge of process and measurement noise covariance matrices. Fuzzy inference system is applied for its simplicity of the approach and the capability of processing imprecise information. The performance of the each filter is evaluated with a confidence degree based on fuzzy logic and covariance-matching technique. The fusion center fuses the local tracks adaptively according to the performance of each filter. The algorithm is implemented in three steps. First, the H_∞ filter is used to process the measurement-vector from each sensor to obtain the local state estimate, thus two H_∞ filters work in parallel. Then fuzzy inference system is employed to assign the confidence degree of each H _∞ filter. Finally, local state estimates are fed to fusion center to obtain the fused state estimate based on the confidence degrees. Simulation results show good performance of the proposed algorithm.