[...]... Instruments and Soft Sensors 2.1 Virtual Instruments In this chapter, the concepts of virtual instruments (VIs) and of soft sensors will be introduced in some detail In particular, we will commence in this section with an introduction to VIs, to focus in the next one on soft sensors, that are the main topic of the book VIs can be considered as a wider class than soft sensors In fact, soft sensors focus... Cascaded Structure for the Soft Sensor 117 6.4 The One-step-ahead Predictor Soft Sensor 127 6.4.1 Refinement of the One-step-ahead Soft Sensor 134 6.5 Conclusions 142 7 Strategies to Improve Soft Sensor Performance 143 7.1 Introduction 143 7.2 Stacked Neural Network Approach for a Sulfur Recovery Unit 144 7.3 Model Aggregation Using Fuzzy Logic for the Estimation... 81 5.1 Introduction 81 5.2 Static Models for the Prediction of NOx Emissions for a Refinery 82 5.3 Linear Dynamic Models for RON Value Estimation in Powerformed Gasoline 87 5.4 Soft Computing Identification Strategies for a Sulfur Recovery Unit 90 5.5 Comparing Different Methods for Inputs and Regressor Selection for a Debutanizer Column 97 5.5.1 Simple Correlation... of neural networks for intelligent sensors and control In recent years a number of books have been published dealing with soft computing and artificial intelligence techniques Some aspects of these fields form the basis of the approaches reported in this book Readers who have no in-depth Soft Sensors in Industrial Applications 5 knowledge of this topic can refer to Haykin (1999), Fortuna et al (2001),... process performance are determined by off-line laboratory analyses Mathematical models of processes designed to estimate relevant process variables can help to reduce the need for measuring devices, improve system reliability and develop tight control policies Plant models devoted to the estimation of plant variables are known either as inferential models, virtual sensors, or soft sensors Soft Sensors. .. models, substituting some physical sensors and using data acquired from some other available ones For their part, VIs are based on software that performs any of the typical actions involved in a measurement and/or control problem, by exploiting available instrumentation, computers and software This action can either involve, or not, modeling capabilities typical of soft sensors VIs are the result of the... microprocessor that controls all actions required to perform measurements, while on the other one, it is more and 18 Soft Sensors for Monitoring and Control of Industrial Processes more common that the instruments communication capabilities allow for the realization of distributed measuring networks where a number of devices cooperate and exchange relevant information, either using a shared standard or custom... proposed in the literature for variable processing is independent component analysis (ICA) It is aimed at making the variables independent, and involves higher order statistics In Lee, Yoo and Lee (2004), ICA is used to process data relative to biological waste water treatment An interesting comparison between ICA and PCA monitoring capabilities is reported 8 Soft Sensors for Monitoring and Control of... study between black-box and hybrid (gray-box) estimation methods for on-line biomass concentration estimation is reported in Janes, Legge and Budman (2002) In Willis et al (1992), an inferential controller, based on a neural network model and a classical PI controller, is used for a penicillin fermentation plant 12 Soft Sensors for Monitoring and Control of Industrial Processes The pulp industry is... devices, allowing the realization of more comprehensive monitoring networks; they can work in parallel with hardware sensors, giving useful information for fault detection tasks, thus allowing the realization of more reliable processes; they can easily be implemented on existing hardware (e.g microcontrollers) and retuned when system parameters change; Soft Sensors in Industrial Applications 3 they allow real-time . role of soft sensors to fault detection and sensor validation configurations is dealt with in Chapter 9. A great strength of Soft Sensors for Monitoring. Professor Fortuna and his colleagues consider soft sensors to be a far narrower concept within the topic of virtual instrumentation, stating that soft sensors