基于人工神经网络电子平台秤故障识别方法  

Study on Identification Method of Artificial Neural Network Electronic Platform Scale Fault

作　　者：刘秀华 项秀明 吕宝超 付建新[1] LIU Xiu-hua;XIANG Xiu-ming;LV Bao-chao;FU Jian-xin(Shougang Jingtang United Iron & Steel Co.,Ltd.,Tangshan 063200 China;Beijing Shougang Automation Information Technology Co.,Ltd.,Beijing 100041 China)

机构地区：[1]首钢京唐钢铁联合有限责任公司,河北唐山063200 [2]北京首钢自动化信息技术有限公司,北京100041

出　　处：《自动化技术与应用》2019年第7期134-136,共3页Techniques of Automation and Applications

摘　　要：为快速实现识别电子平台秤故障,保证电子平台秤正常的工作状态,此方法针对电子平台秤四个数字传感器的输出码值进行研究,通过采集不同状态下个传感器的输出码值,然后使用Matlab软件中的BP神经网络的建立识别模型。结果表明,通过四种状态下传感器码值的输出值建立4(输入节点)-10(隐层节点)-2(输出节点)的BP人工神经网络模型,该神经网络模型的预测准确率可达到100%。因此认为,人工神经网络算法可为电子平台秤的故障识别提供可靠的技术支撑。In order to identify electronic platform scale failure quickly and ensure that the electronic platform scale normal working conditions, the method is used to study the output code value of four digital sensors of electronic platform scale. By collecting the output code value of the sensor in different states, the BP neural network in Matlab is used to establish the identification model. By four kinds of condition sensor output value of the code value set up 4(input node) to 10(hidden layer nodes)-2(output node) of BP artificial neural network model, the prediction accuracy of neural network model can reach100%. Therefore, artificial neural network algorithm can provide reliable technical support for fault identification of electronic platform scale.

关 键 词：电子平台秤 数字传感器 人工神经网络 故障识别 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]