A Multi-sensor Fusion Algorithm with Feedback Based on Fuzzy C-Means and Maximum Entropy Principle

Aiming at the disadvantages of high computation overhead and bad extensibility in matrix weighted fusion method, a multi-sensor fusion algorithm with feedback based on fuzzy c-means (FCM) clustering and maximun entropy principle (MEP) is proposed in this paper. This algorithm combined FCM and MEP to calculate fusion matrix weight of local state estimates considering every component of state vector integratedly. What´s more, this algoritm has a good real-time performance due to less matrix computation and good extensibility which show it can directly be applied into tracking system comprising more than two sensors. It is proved by experiments and results that the tracking accuracy of fusion estimate is higher than that of matrix weighted fusion method.