This thesis develops a systematic, data-based dynamic modeling framework for industrial processes in keeping with the slowness principle. Using said framework as a point of departure, it then proposes novel strategies for dealing with control monitoring and quality prediction problems in industrial production contexts. The thesis reveals the slowly varying nature of industrial production processes under feedback control, and integrates it with process data analytics to offer powerful prior knowledge that gives rise to statistical methods tailored to industrial data. It addresses several issues of immediate interest in industrial practice, including process monitoring, control performance assessment and diagnosis, monitoring system design, and product quality prediction. In particular, it proposes a holistic and pragmatic design framework for industrial monitoring systems, which delivers effective elimination of false alarms, as well as intelligent self-running by fully utilizing the information underlying the data. One of the strengths of this thesis is its integration of insights from statistics, machine learning, control theory and engineering to provide a new scheme for industrial process modeling in the era of big data.

Data-driven dynamical systems is a burgeoning field?it connects how measurements of nonlinear dynamical systems and/or complex systems can be used with well-established methods in dynamical systems theory. This is a critically important new direction because the governing equations of many problems under consideration by practitioners in various scientific fields are not typically known. Thus, using data alone to help derive, in an optimal sense, the best dynamical system representation of a given application allows for important new insights. The recently developed dynamic mode decomposition (DMD) is an innovative tool for integrating data with dynamical systems theory. The DMD has deep connections with traditional dynamical systems theory and many recent innovations in compressed sensing and machine learning. Dynamic Mode Decomposition: Data-Driven Modeling of Complex Systems, the first book to address the DMD algorithm, presents a pedagogical and comprehensive approach to all aspects of DMD currently developed or under development; blends theoretical development, example codes, and applications to showcase the theory and its many innovations and uses; highlights the numerous innovations around the DMD algorithm and demonstrates its efficacy using example problems from engineering and the physical and biological sciences; and provides extensive MATLAB code, data for intuitive examples of key methods, and graphical presentations.

The International Conference on Computational Science (ICCS 2004) held in Krak ́ ow, Poland, June 6–9, 2004, was a follow-up to the highly successful ICCS 2003 held at two locations, in Melbourne, Australia and St. Petersburg, Russia; ICCS 2002 in Amsterdam, The Netherlands; and ICCS 2001 in San Francisco, USA. As computational science is still evolving in its quest for subjects of inves- gation and e?cient methods, ICCS 2004 was devised as a forum for scientists from mathematics and computer science, as the basic computing disciplines and application areas, interested in advanced computational methods for physics, chemistry, life sciences, engineering, arts and humanities, as well as computer system vendors and software developers. The main objective of this conference was to discuss problems and solutions in all areas, to identify new issues, to shape future directions of research, and to help users apply various advanced computational techniques. The event harvested recent developments in com- tationalgridsandnextgenerationcomputingsystems,tools,advancednumerical methods, data-driven systems, and novel application ?elds, such as complex - stems, ?nance, econo-physics and population evolution.

Zhiwen Chen aims to develop advanced fault detection (FD) methods for the monitoring of industrial processes. With the ever increasing demands on reliability and safety in industrial processes, fault detection has become an important issue. Although the model-based fault detection theory has been well studied in the past decades, its applications are limited to large-scale industrial processes because it is difficult to build accurate models. Furthermore, motivated by the limitations of existing data-driven FD methods, novel canonical correlation analysis (CCA) and projection-based methods are proposed from the perspectives of process input and output data, less engineering effort and wide application scope. For performance evaluation of FD methods, a new index is also developed.

A typical design procedure for model predictive control or control performance monitoring consists of: 1. identification of a parametric or nonparametric model; 2. derivation of the output predictor from the model; 3. design of the control law or calculation of performance indices according to the predictor. Both design problems need an explicit model form and both require this three-step design procedure. Can this design procedure be simplified? Can an explicit model be avoided? With these questions in mind, the authors eliminate the first and second step of the above design procedure, a “data-driven” approach in the sense that no traditional parametric models are used; hence, the intermediate subspace matrices, which are obtained from the process data and otherwise identified as a first step in the subspace identification methods, are used directly for the designs. Without using an explicit model, the design procedure is simplified and the modelling error caused by parameterization is eliminated.

Modeling and Control of Batch Processes presents state-of-the-art techniques ranging from mechanistic to data-driven models. These methods are specifically tailored to handle issues pertinent to batch processes, such as nonlinear dynamics and lack of online quality measurements. In particular, the book proposes: a novel batch control design with well characterized feasibility properties; a modeling approach that unites multi-model and partial least squares techniques; a generalization of the subspace identification approach for batch processes; and applications to several detailed case studies, ranging from a complex simulation test bed to industrial data. The book’s proposed methodology employs statistical tools, such as partial least squares and subspace identification, and couples them with notions from state-space-based models to provide solutions to the quality control problem for batch processes. Practical implementation issues are discussed to help readers understand the application of the methods in greater depth. The book includes numerous comments and remarks providing insight and fundamental understanding into the modeling and control of batch processes. Modeling and Control of Batch Processes includes many detailed examples of industrial relevance that can be tailored by process control engineers or researchers to a specific application. The book is also of interest to graduate students studying control systems, as it contains new research topics and references to significant recent work. Advances in Industrial Control reports and encourages the transfer of technology in control engineering. The rapid development of control technology has an impact on all areas of the control discipline. The series offers an opportunity for researchers to present an extended exposition of new work in all aspects of industrial control.