摘要
针对燃气管道阀门故障诊断存在的诊断准确率低,鲁棒性差和容易陷入过拟合等问题,结合深度学习理论,基于谷歌人工智能学习系统Keras,构建多层感知器MLP神经网络模型,用于预测阀门故障程度。选取阀门故障中的8种特征参数作为模型的原始输入量,经过多层感知器的特征提取、参数重构、Adam优化、Softmax分类,并加入Dropout模块避免过度拟合,最终得到具有较高预测精度的多层感知器模型。将得到的多层感知器模型应用在实验室的燃气管道阀门故障诊断系统中,结果表明,这种模型具有更高的准确性和鲁棒性。
Aiming at the problems of low diagnostic accuracy,poor robustness and easy to fall into over fitting in the fault diagnosis of gas pipeline valves,combining with the deep learning theory,a MLP neural network model based on Google artificial intelligence learning system Keras is proposed to predict the fault degree of valves.Eight characteristic parameters of valve fault are selected as the original input of the model.After feature extraction,parame⁃ter reconstruction,Adam optimization,Softmax classification of multi-layer perceptron,and adding dropout module to avoid over fitting,the multi-layer perceptron model with high prediction accu⁃racy is finally obtained.The multi-layer perceptron model is ap⁃plied to the gas pipeline valve fault diagnosis system in the labora⁃tory.The results show that this method has higher accuracy and ro⁃bustness.
作者
王新颖
张瑞程
张惠然
赵斌
黄旭安
WANG Xin-ying;ZHANG Rui-cheng;ZHANG Hui-ran;ZHAO Bin;HUANG Xu-an(School of Environment and Safety Engineering,Changzhou University,Jiangsu Changzhou 213164,China)
出处
《消防科学与技术》
CAS
北大核心
2020年第4期541-546,共6页
Fire Science and Technology
基金
常州市科技项目“城市地下燃气管网信息化管理与应急决策支持系统”(CZ20170017)。