神经网络在多传感器信息集成与融合中的应用  

Application of Neural Network in Multisensor Information Integration and Fusion

作　　者：周超超[1] 张永志[1] 程胜文[1] 

机构地区：[1]武汉理工大学东院机电工程学院,湖北武汉430070

出　　处：《仪表技术与传感器》2002年第11期49-50,54,共3页Instrument Technique and Sensor

摘　　要：多传感器信息集成与融合在处理信息中呈现出较好的实用性和优越性。介绍了有关多传感器信息集成与融合方面的基本知识 ,分析了传统的信息融合与运用神经网络实现多传感器信息融合的差异、实质 ,并给出了神经网络实现多传感器信息集成与融合的步骤。通过自动化程度很高的智能机器人对外界信息的集成与融合 ,用实验证明了神经网络比传统方法更能准确。A new technology emerges recently to the multisensor information integration and fusion(in short of MIIF),which takes on better practicality and superiority.The paper introduces the basic knowledge of MIIF,compares the differentia and substance of traditional & neural network-realizing MIIF,compares the differnentia and substance of traditional & neural network-realizing MIIF,and educes the procedures of multisensor fusion by means of neural network.Finally the experiment proved that the neural network could describe the attained information of multisensor more accurately and reliably than the traditional methods by high automated intelligent robot.

关 键 词：多传感器 信息集成 信息融合 神经网络 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置] TP183[自动化与计算机技术—控制科学与工程]