多目标数据关联的神经网络解算  

A Neural Network Method Applied to Multi-target Data Association Processing

作　　者：逯宏亮[1] 李伟仁[1] 

机构地区：[1]空军工程大学工程学院,陕西西安710038

出　　处：《系统仿真学报》2004年第7期1536-1538,共3页Journal of System Simulation

摘　　要：从组合优化的角度,建立了密集回波条件下多目标数据关联的数学模型,并提出连续型Hopfield神经网络解算方案,对结果进行了仿真。通过与联合概率数据关联(JPDA)方法的对比仿真,表明了该算法的有效性和快速性。From the point of the combination optimization view, a data association model is established under the dense multi-target environments, and then a continuous Hopfield neural network (CHNN) is designed and applied to the processing. The simulation result shows that this algorithm is more effective and efficient compared with the joint probability data association (JPDA) method.

关 键 词：多目标跟踪 数据关联 神经网络 算法 

分 类 号：TN953[电子电信—信号与信息处理]