摘要
针对燃气管道闸阀故障诊断存在的诊断准确率低、运算过程复杂等问题,结合深度学习理论,基于人工智能学习软件TensorFlow,自行设计张量运算,构建深度学习神经网络模型,用于预测闸阀发生故障的严重程度。选择了闸阀在故障预测系统中的9种参数,经过标准化处理,卷积神经网络进行特征提取、运算、故障评估等分析,最终设计得到了具有高精度预测的卷积神经网络模型。
Aiming at the problems of low diagnostic accuracy and complex operation processes in the fault diagnosis of gas pipeline gate valve,combined with the deep learning theory,the fault diagnosis method of gas pipeline gate valve is proposed,based on tensorflow,an artificial intelligence learning software,tensor operation is designed and a deep learning neural network model is constructed to predict the severity of gate valve failure.Nine parameters of gate valve in fault prediction system are selected and standardized,convolution neural network is used for feature extraction,operation,fault evaluation and other analysis,and finally a convolution neural network model with high precision prediction is designed.
作者
黄旭安
王新颖
林振源
胡磊磊
刘岚
HUANG Xu′an;WANG Xinying;LIN Zhenyuan;HU Leilei;LIU Lan(School of Environment and Safety Engineering, Changzhou University Changzhou, Jiangsu 213164)
出处
《工业安全与环保》
2021年第10期39-43,共5页
Industrial Safety and Environmental Protection
基金
江苏省研究生科研与实践创新计划项目(KYCX21-2883)。
