报告名称： | 基于慢特征分析和神经网络的非线性软件传感器 | |
Report Name: | Nonlinear Software Sensors based on Slow Feature Analysis and Neural Networks | |
报告人姓名： | Dr Jie Zhang |
|
Name: | Dr Jie Zhang | |
报告时间： | 2024-07-19 16:00 | |
报告地点： | 会议中心二层报告厅 | |
邀请团队或课题组名称 | 安全分析与智能决策研究室 | |
报告简介（200-400字）： 在过去的二十年中，软件传感器（常称为软传感器）已经成为监测、控制和优化工艺过程中获取关键过程变量的有价值的补充方式，有时甚至是替代传统手段的选择。本次演讲介绍了将慢特征分析（Slow Feature Analysis，简称SFA）与神经网络（Neural Networks，简称NN）集成用于软传感器开发。首先，将动态线性SFA应用于易于测量的过程变量数据中。然后，选择主要的慢特征作为神经网络的输入，用于预测难以测量的产品质量变量。SFA能够通过提取缓慢变化的潜在变量（即慢特征）来捕捉工业过程的潜在动态特性。我们提出了使用屏幕图（scree plot）来选择主要的慢特征。神经网络被用来处理许多实际工业过程中存在的非线性问题。我们将提出的方法与慢特征回归、偏最小二乘回归、传统前馈神经网络以及在神经网络之前使用主成分分析进行比较，评估了该方法在三个真实工业过程中的有效性。在两个案例研究中，提出的SFA-NN方法在泛化性能方面表现最好。 |
Introduction to the Report (200-300 words): In the last two decades Software Sensors, often referred to as Soft Sensors, have established themselves as a valuable addition, sometimes alternative, to traditional means for the acquisition of critical process variables for process monitoring, process control and process optimisation. This talk presents integrating slow feature analysis (SFA) with neural networks (SFA-NN) for soft sensor development. Dynamic linear SFA is applied to the easy to measure process variable data. Then the dominant slow features are selected as the inputs of a neural network to predict the difficult to measure product quality variables. SFA can capture underlying dynamics of industrial processes through the extraction of slowly varying latent variables, known as slow features. Selection of dominant slow features using scree plot is proposed. Neural networks are utilised to cope with nonlinearities present in many real industrial processes. The effectiveness of the proposed method is evaluated on three real industrial processes and is compared with slow feature regression, partial least square regression, traditional feedforward neural networks, and using principal component analysis prior to a neural network. The proposed SFA-NN gives the best generalisation performance among these techniques in both case studies. |
报告人简介（200-400字）：Dr Jie Zhang 1991 年获得伦敦城市大学控制工程博士学位。自 1991 年以来，他一直在英国纽卡斯尔大学工程学院工作，现任过程系统工程教授。他的研究兴趣是过程系统工程的一般领域，包括过程建模、批量过程控制、过程监控和计算智能。他在国际期刊、书籍和会议论文集上发表了 400 多篇论文。他指导了 30 多名博士生和 100 多名硕士生完成学业。他是IEEEsenior member，同时担任包括Neurocomputing、Network: Computation in Neural Systems、Processes和PLOS ONE等许多期刊的编委。 |
Biography (200-300 words): Dr Jie Zhang received his PhD in Control Engineering from City University, London, in 1991. He has been with the School of Engineering, Newcastle University, UK, since 1991 and is Reader in Process Systems Engineering. His research interests are in the general areas of process system engineering including process modelling, batch process control, process monitoring, and computational intelligence. He has published over 400 papers in international journals, books, and conference proceedings. He has supervised over 30 PhD students and over 100 master students to completion. He is a senior member of IEEE. He is on the Editorial Boards of a number of journals including Neurocomputing, Network: Computation in Neural Systems, Processes, and PLOS ONE. |