Multi-sensor Information Fusion, With State Constraints

With the need of modern war and the development of science and technology, information fusion is given wide attention; as a new subject it has developed fast in recent years. Position fusion level is one of the most important and widely used level in the multi-sensor information fusion system. In position fusion level an important question is how to decide whether measurements or tracks coming from different sensors represent the same target, as is called the measurement or track-to-track association problem. If so, the next problem is how to combine the measurement or the track estimations together to get the global optimal state estimation.The problem of state estimation in the information fusion system is studied in this paper after reviewing the history and what is going on about the information fusion. Under the assumption of the completion of target detection, data calibration and association, the previous fusion structure models and algorithms are mended, and the new ones are given. Based on the single sensor Kalman filter algorithm and the distributed multi-sensor fusion algorithm without state constraints, the distributed multi-sensor fusion algorithm with state constraints is proposed utilizing the Kalman filter with constraints. The optimal weighted fusion algorithms with state constraints are derived based upon the Kalman filter with constraints and the optimal weighted idea. A new state estimation fusion algorithm is got in the multi-sensor information fusion system through the improvement of the information processing model in the hybrid fusion architecture. Furthermore this paper gives the brief theory analysis of the proposed fusion algorithms. The simulations indicate the availability of these algorithms which have some value for solving issue on target tracking problem in multi-sensor information fusion system. In the end the main work of this paper is summarized and further work is proposed at the aspect of state estimation in multi-sensor information fusion system.