Fuzzy Modeling For Control

Rule-based fuzzy modeling has been recognised as a powerful technique for the modeling of partly-known nonlinear systems. Fuzzy models can effectively integrate information from different sources, such as physical laws, empirical models, measurements and heuristics. Application areas of fuzzy models include prediction, decision support, system analysis, control design, etc. Fuzzy Modeling for Control addresses fuzzy modeling from the systems and control engineering points of view. It focuses on the selection of appropriate model structures, on the acquisition of dynamic fuzzy models from process measurements (fuzzy identification), and on the design of nonlinear controllers based on fuzzy models. To automatically generate fuzzy models from measurements, a comprehensive methodology is developed which employs fuzzy clustering techniques to partition the available data into subsets characterized by locally linear behaviour. The relationships between the presented identification method and linear regression are exploited, allowing for the combination of fuzzy logic techniques with standard system identification tools. Attention is paid to the trade-off between the accuracy and transparency of the obtained fuzzy models. Control design based on a fuzzy model of a nonlinear dynamic process is addressed, using the concepts of model-based predictive control and internal model control with an inverted fuzzy model. To this end, methods to exactly invert specific types of fuzzy models are presented. In the context of predictive control, branch-and-bound optimization is applied. The main features of the presented techniques are illustrated by means of simple examples. In addition, three real-world applications are described. Finally, software tools for building fuzzy models from measurements are available from the author.

In the last ten years, a true explosion of investigations into fuzzy modeling and its applications in control, diagnostics, decision making, optimization, pattern recognition, robotics, etc. has been observed. The attraction of fuzzy modeling results from its intelligibility and the high effectiveness of the models obtained. Owing to this the modeling can be applied for the solution of problems which could not be solved till now with any known conventional methods. The book provides the reader with an advanced introduction to the problems of fuzzy modeling and to one of its most important applications: fuzzy control. It is based on the latest and most significant knowledge of the subject and can be used not only by control specialists but also by specialists working in any field requiring plant modeling, process modeling, and systems modeling, e.g. economics, business, medicine, agriculture,and meteorology.

This book is focused on mathematical analysis and rigorous design methods for fuzzy control systems based on Takagi-Sugeno fuzzy models, sometimes called Takagi-Sugeno-Kang models. The author presents a rather general analytical theory of exact fuzzy modeling and control of continuous and discrete-time dynamical systems. Main attention is paid to usability of the results for the control and computer engineering community and therefore simple and easy knowledge-bases for linguistic interpretation have been used. The approach is based on the author’s theorems concerning equivalence between widely used Takagi-Sugeno systems and some class of multivariate polynomials. It combines the advantages of fuzzy system theory and classical control theory. Classical control theory can be applied to modeling of dynamical plants and the controllers. They are all equivalent to the set of Takagi-Sugeno type fuzzy rules. The approach combines the best of fuzzy and conventional control theory. It enables linguistic interpretability (also called transparency) of both the plant model and the controller. In the case of linear systems and some class of nonlinear systems, engineers can in many cases directly apply well-known classical tools from the control theory both for analysis, and the design of closed-loop fuzzy control systems. Therefore the main objective of the book is to establish comprehensive and unified analytical foundations for fuzzy modeling using the Takagi-Sugeno rule scheme and their applications for fuzzy control, identification of some class of nonlinear dynamical processes and classification problem solver design.

Fuzzy Modeling and Control: Methods, Applications and Research opens by recommending a new fuzzy RANSAC algorithm based on the reinforcement learning concept to improve modeling performance under the outlier noise. The authors also propose a novel methodology for online modeling of multivariable Hammerstein evolving fuzzy models with minimum realization in state space from experimental data. Results characterized by strongly coupled nonlinearities demonstrate the computational efficiency of the proposed methodology. Later, two types of neural networks are applied to find the approximate solutions of the fully fuzzy nonlinear system, and a superior gradient descent algorithm is proposed in order to train the neural networks. Lastly, the authors propose a novel online evolving fuzzy Takagi-Sugeno state-space model identification approach for nonlinear multivariable systems. To circumvent "the curse of dimensionality", the algorithm uses tools for monitoring the quality of the existing clusters. (Novinka)

This collection compiles the seminal contributions of Michio Sugeno on fuzzy systems and technologies. Fuzzy Modeling & Control: Selected Works of Sugeno serves as a singular resource that provides a clear, comprehensive treatment of fuzzy control systems. The book comprises two parts fuzzy system identification and modeling systems control Each part outlines the fundamentals of fuzzy logic and covers essential material for understanding the mathematical and modeling details in Sugeno's works. Introductory chapters include extended summaries of each paper or group of papers, suggesting where the theories discussed might be useful in application.

This book offers a thorough introduction to the field of fuzzy logic with complete coverage of both relevant theory and applications. With its comprehensive presentation of fuzzy logic as well as coverage of both fuzzy control and modeling, this text is destined to become the classic primer in this quickly developing field.

Fuzzy logic methodology has proven effective in dealing with complex nonlinear systems containing uncertainties that are otherwise difficult to model. Technology based on this methodology is applicable to many real-world problems, especially in the area of consumer products. This book presents the first comprehensive, unified treatment of fuzzy modeling and fuzzy control, providing tools for the control of complex nonlinear systems. Coverage includes model complexity, model precision, and computing time. This is an excellent reference for electrical, computer, chemical, industrial, civil, manufacturing, mechanical and aeronautical engineers, and also useful for graduate courses in electrical engineering, computer engineering, and computer science.

Decision making and control are two fields with distinct methods for solving problems, and yet they are closely related. This book bridges the gap between decision making and control in the field of fuzzy decisions and fuzzy control, and discusses various ways in which fuzzy decision making methods can be applied to systems modeling and control.Fuzzy decision making is a powerful paradigm for dealing with human expert knowledge when one is designing fuzzy model-based controllers. The combination of fuzzy decision making and fuzzy control in this book can lead to novel control schemes that improve the existing controllers in various ways. The following applications of fuzzy decision making methods for designing control systems are considered: OCo Fuzzy decision making for enhancing fuzzy modeling. The values of important parameters in fuzzy modeling algorithms are selected by using fuzzy decision making.OCo Fuzzy decision making for designing signal-based fuzzy controllers. The controller mappings and the defuzzification steps can be obtained by decision making methods.OCo Fuzzy design and performance specifications in model-based control. Fuzzy constraints and fuzzy goals are used.OCo Design of model-based controllers combined with fuzzy decision modules. Human operator experience is incorporated for the performance specification in model-based control.The advantages of bringing together fuzzy control and fuzzy decision making are shown with multiple examples from real and simulated control systems."