ABSTRACT
Artificial neural networks (ANNs) are powerful tools to map nonlinear relationships between variables with limited, incomplete, nonintegrated, uncertain, noisy, dynamic, multidimensional, and nonlinear databases owing to their excellent information-processing capabilities such as nonlinearity, high parallelism, robustness, and failure tolerance. This chapter explains the principles of the ANN approach and provides an overview of the most important investigations on the application of this paradigm in drying technology. An ANN is a group of interconnected artificial neurons, interacting with one another in a concerted manner. Adaptive neuro-fuzzy inference system (ANFIS) is a hybrid intelligent system, which generates fuzzy rules from a given input-output data set by implementing the Takagi-Sugeno fuzzy inference system. ANFIS is a six-layer generalized network with supervised learning, composed of input, fuzzification, rule, normalization, and defuzzification layers as well as a single summation node. It can be trained using a hybrid learning algorithm through integrating a backpropagation (BP) algorithm with the least squares method.
