神经网络自学习技术在传感器数据观测中的应用

摘要

传感器作为自动化领域的核心组成部件,其数据对系统下一步的行为决策起决定性作用.因此正确的观测传感器实时数据并分析其是否存在异常有着十分重要的意义.鉴于当前故障诊断、异常检测等方法存在的局限性,提出了一种基于神经网络自学习的传感器数据观测方法.在此基础上,以BP神经网络和RBF神经网络为例,通过一个实际的对象,验证了该文提出方法的有效性,同时分析讨论了2种神经网络在该过程中的优劣性.%As a core component in automation area,sensors have a direct impact for system behaviors or strategies.Thus,there is a very important significance to observe their data and analysis the abnormal behaviors.Because of the boundedness of the current methods for fault diagnosis and anomaly detection,sensors' data observation approach based on neural network is proposed in this paper.On the basis of it,as examples of BP neural network and RBF neural network,the effectiveness of the proposed approach has been verified by an actual object.At last,the advantages and disadvantages between the two neural network have been discussed.