ABSTRACT
As shown in the previous chapters, most development environments for designing fuzzy systems implement them with a high-level programming lan guage. Software implementations of fuzzy systems are easily obtained and pro vide a high flexibility since they usually support fuzzy systems with an arbitrary number of rules, without any limitations concerning the number and type of membership functions, and with a wide range of inference mechanisms. On the other hand, software approaches are not adequate for applications de manding fuzzy systems with a small size, low power consumption, and high in ference speed. To verify the requirements imposed by these kinds of applica tions, hardware approaches have to be adopted. This issue is addressed in the first section of this chapter by briefly describing the different ways of imple menting fuzzy systems depending on the application. The rest of the chapter re views the different hardware approaches, also known as fuzzy hardware. There are two basic procedures for developing fuzzy hardware. One approach is to employ general-purpose microprocessors, occasionally expanded with new in structions or new circuitry. The other approach is to design dedicated hard ware, that is, application-specific integrated circuits (ASICs), optimized for fuzzy logic-based inference systems. The last sections of the chapter focus on the second procedure. Several strategies to implement fuzzy systems with dedicated hardware are analyzed and the fundamental concepts concerning the design and fabrication of integrated circuits (ICs) are discussed. This material serves as an introduction to the following chapters where microelectronic realizations of fuzzy systems with different design techniques are described in greater detail.
