This book discusses the problems of complexity in industrial data, including the problems of data sources, causes and types of data uncertainty, and methods of data preparation for further reasoning in engineering practice. Each data source has its own specificity, and a characteristic property of industrial data is its high degree of uncertainty. The book also explores a wide spectrum of soft modeling methods with illustrations pertaining to specific cases from diverse industrial processes. In soft modeling the physical nature of phenomena may not be known and may not be taken into consideration. Soft models usually employ simplified mathematical equations derived directly from the data obtained as observations or measurements of the given system. Although soft models may not explain the nature of the phenomenon or system under study, they usually point to its significant features or properties.

Discusses the problems of industrial data and it overall complexity, as well as methods of dealing with preparing this type of data for reasoning in the engineering practice Presents a wide spectrum of soft modeling methods, with illustrations pertaining to specific cases, coming from diverse kinds of industrial processes Includes recent research in soft modeling in industrial manufacturing Offers a variety of applications and diverse methodologies relating to with a wide range of soft modeling methods Provides an analysis of a specific practical problem in each chapter, in addition to a detailed discussion of the proposed modeling method

Inhalt

Data and modeling in industrial manufacturing.- From data to reasoning.- Data preprocessing in industrial manufacturing.- Tool condition monitoring in metal cutting.- Assessment of selected tools used for knowledge extraction in industrial manufacturing.- Application of data mining tools in shrink sleeve labels converting process.- Study of thickness variability of the oorboard surface layer.- Applying statistical methods with imprecise data to quality control in cheese manufacturing.- Monitoring series of dependent observations using the sXWAM control chart for residuals.- Diagnosis of out-of-control signals in complex manufacturing processes.