检索结果-南通市图书馆

检索条件"主题词=dynamic soft sensor"

共 2 条记录，以下是1-10 订阅

全选清除本页清除全部题录导出标记到"检索档案"

详细简洁

排序：

Semi-supervised dynamic soft sensor Based on Complementary Ensemble Empirical Mode Decomposition and Gated Recurrent Units

Semi-supervised Dynamic Soft Sensor Based on Complementary E...

引用

第31届中国过程控制会议（CPCC2020）

作者： GUO Runyuan LIU Han XIE Guo ZHANG Youmin School of Automation and Information Engineering Xi’an University of Technology

[Objectives] Noise, redundancy and dynamic characteristics in complex industrial process data have been regarded as the key factors that affect the prediction accuracy of data-driven soft sensor models. In this paper, a semi-supervised dynamic soft sensor method is proposed to capture the dynamic characteristics of variable data while removing the noise and redundancy of the data to establish an accurate dynamic soft sensor prediction model. [Methods] First, auxiliary variable signals are decomposed effectively by complementary ensemble empirical mode decomposition(CEEMD), and the noise is eliminated based on the intrinsic mode functions(IMFs) obtained by the decomposition. Then, using the isometric feature mapping(Isomap) method to perform feature transformation of the remaining intrinsic mode functions, the variable data are reconstructed to remove the internal redundancy and noise. Finally, the preprocessed and serialized data are used to build a deep learning model based on a semi-supervised deep gated recurrent unit network aided multilayer perceptron(SSDGRU-MLP). To ensure effective training of the proposed SSDGRU-MLP model, callback functions and dropout technology are adopted. After the above offline procedures, a semi-supervised dynamic soft sensor prediction model can be obtained. Then, by importing the new serialized dataset into the established soft sensor model, the predicted values of the primary variable can be obtained through propagation over time and network depth. Since power plant air preheater is a dynamic system with complex and variable operating conditions, the semi-supervised dynamic soft sensor method proposed in this paper is used to predict rotor thermal deformation in the air preheater. So, the effectiveness and superiority of the new dynamic soft sensor are verified by this real industrial case. In our experiments, four auxiliary variables(gas temperature of the flue inlet, gas temperature of the flue outlet, air temperature of the air

关键词： dynamic soft sensor complementary ensemble empirical mode decomposition deep learning gated recurrent unit semi-supervised and dynamic modeling isometric feature mapping

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

Development of dynamic soft Senors based on Bayesian Semi-supervised Nonlinear HMM

Development of Dynamic Soft Senors based on Bayesian Semi-su...

引用

第35届中国过程控制会议

作者： Xu Li Hongjian Yu Weiming Shao College of New Energy China University of Petroleum (East China)

Real-time monitoring of product quality-related key variables is crucial for ensuring product quality and optimization of control and decision making in modern process industries. Hidden Markov model(HMM) has proven effective for monitoring key variables in dynamic and multimode industrial processes. However, the majority of HMMs assume a linear relationship between input and output variables, which faces great challenges in the application of strong nonlinear industrial ***, a novel Bayesian semi-supervised nonlinear hidden Markov model(BSsNHMM) is proposed. In the BSsNHMM,the dependence relationship between input and output is learned by the neural network, and the learning of dynamic and nonlinear features is accomplished under a unified framework. Besides, a variational inference-and gradient ascent-based training algorithm is developed to learn the post distribution of variables in HMM and the estimated parameter of neural network. The performance of the BSsNHMM is evaluated using both numerical and real industrial cases for virtual sensing, demonstrating its enormous potential for soft sensing applications in complex process industries.

关键词： Semi-supervised learning Nonlinear HMM Variational inference dynamic soft sensor

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

全选清除本页清除全部题录导出标记到“检索档案”

共1页 << < 1 > >>

回到顶部

执行限定条件

欢迎您,

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

在线全文

在线全文

请选择保存的检索档案：

请选择收藏分类：

通借通还

欢迎您,

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

在线全文

在线全文

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：