In view of the problems that sole guide system cannot meet the modern battlefield demand and data update of multi-sensor is not synchronized, a new MIMU/MMW/IR (miniature inertial measurement unit/ millimeterwave/infrared) composite guidance system was built and the time synchronization and space match problems of multi-sensor asynchronous information fusion in this system were studied. A novel adaptive UKF (unscented Kalman filter) algorithm was presented. The statistics of status model error was built by prediction error in this algorithm; contribution of status model information to status parameter estimation was adjusted by adaptive factor. So the influence of status model noise is effectively regulated. The proposed algorithm was applied to the MIMU/MMW/IR compound guidance system and the algorithm performance was tested. The simulation results show that the adaptive UKF is better than the standard UKF and extended Kalman filter, and it can improve the positioning precision effectively.