Research on Track-to-track Fusion Under Complication Environment

In actual multisensor application systems,some sensors or lines of communication would be cause to fault, which sequentially lead to lacking data or appearing outlier in center fusion,and the center fusion algorithm can't be executed ideally.The method of multisensor data fusion is apposed,which can real-time eliminate the outlier and adapt to data lacking.In this method,the track of multisensor is clustered using Fuzzy C-means firstly,the outliers are detected and eliminated based on the compact and the membership.Finally forecasted estimation replaces outliers and lacking data to fusion.Simulation results showed that the algorithm could resolve the problem of outlier eliminating and data lacking efficiently.