ABSTRACT
Ever since ANN was presented for the first time by Bernard Widrow from Stanford University in 1950s, it has been a constant challenge for researchers to find optimal ANN-based solutions to design, manufacture, develop, and operate new generations of industrial systems as efficiently, reliably, and durably as possible. Getting enough information about the system that is to be modeled is the first step in system identification and modeling process. Besides, a clear statement of the modeling objectives is necessary for making an efficient model. Industrial systems may be modeled for condition monitoring, fault detection and diagnosis, sensor validation, system identification or design, and optimization of control systems [209].
