A Target Location Method Based on Swarm Probability Fusion

The existing target location methods either depend on the sensors' high performance or the communication bandwidth, which are not suitable to directly apply to the effective fusion and target recognition of observation information of the large-scale heterogeneous swarm. In this paper, a target location method based on swarm probability fusion is proposed, which can represent the sensor's data in the heterogeneous swarms as the probability distribution map, and then carry out image recognition processing after normalized distributed fusion to obtain target information. The effectiveness of the method is verified by the effectiveness tests, the singlenode's direction finding accuracy tests, and the swarm nodes' number tests. Preliminary analysis shows that single-node's direction finding accuracy has a significant effect on the swarm positioning accuracy and the convergence speed, while the number of nodes has a small effect on positioning accuracy and mainly affects the convergence speed.