Multi-sensor distributed fusion based on CFSFDP clustering algorithm

Abstract

Multi-sensor and multi-target distributed fusion has a big computational burden. A multi-sensor distributed fusion algorithm based on clustering by fast search and find of density peaks (CFSFDP) is proposed. The local target tracks which come from the multi-sensor are divided into multiple categories. The obtained categories number is the target number. The corresponding cluster centers are the fusion states of the targets. The least square fitting algorithm is used to smooth the target's fusion result. The simulation results show that the proposed algorithm has a higher robustness, compared with traditional Covariance Intersection (CI) fusion and local tracks of single sensor.