基于人工智能的多传感器多目标无源定位仿真  

Passive Location Simulation of Multi-Sensor and Multi-Target Based on Artificial Intelligence

作　　者：苏虹 SU Hong(Zhengzhou University of Light Industry,Zhengzhou Henan 450000,China)

机构地区：[1]郑州轻工业大学,河南郑州450000

出　　处：《计算机仿真》2020年第9期399-403,共5页Computer Simulation

摘　　要：传统无源定位方法容易受到复杂环境干扰,产生大量虚假定位点,导致定位精准度降低、定位时间过长,因此提出一种基于人工智能的多传感多目标无源定位方法。通过对传感器获得的方位数据做滤波处理和基于聚类复合弹性神经网络的关联优化,以获得不同时间段准确的目标角度信息;在传感目标均可被观测以及传感半径足够大的条件下,分别建立目标动态、目标测量和移动传感器运动状态模型;根据模型分析定位原理,利用关联优化数据实现对目标的交叉定位,通过数据融合,完成多传感器多目标的无源定位。实验结果表明,所提方法能够准确、且快速的进行多传感器多目标无源定位,具有重要的理论与实际应用价值。Traditionally,the passive location method is interfered easily by complex environment,thus generating false location points.Therefore,a multi-sensor and multi-target passive location method based on artificial intelli⁃gence was proposed.By filtering the azimuth data obtained by sensor and optimizing the association based on the clus⁃tering composite elastic neural network,we got the accurate angle information of target in different time periods.If the sensing target could be observed and the sensing radius was large enough,the dynamic model of target,the measure⁃ment model of target and the motion model of mobile sensor were built respectively.According to the model,we ana⁃lyzed the positioning theory and used the correlation optimization data to complete the cross location for target.Final⁃ly,the passive location of multi-sensor and multi-target was completed by data fusion.Simulation results show that the proposed method can perform multi-sensor and multi-target passive location accurately and quickly,so it has im⁃portant theoretical and practical value.

关 键 词：人工智能 多传感器 多目标 无源定位 神经网络 

分 类 号：TP912[自动化与计算机技术]