Nonlinear neuro-fuzzy prediction: methodology, design and applications

This paper presents the use of neuro-fuzzy networks as means of implementing algorithms suitable for nonlinear black-box modeling and prediction. Artificial neural networks and fuzzy logic systems have recently been emerged as attractive tools for engineering applications. One area of particular importance is the design of networks capable of modeling and prediction of the behavior of systems that involve complex, multi-variable processes with time-variant parameters. This paper shows that neuro-fuzzy networks lend themselves well to nonlinear black-box modeling and prediction. To illustrate the applicability of the neuro-fuzzy networks, a case study involving electric arc furnace is presented here. Electric arc furnaces represent complex, nonlinear loads with stochastic behavior and as such prediction and modeling of furnace characteristics are demanding tasks. Successful implementations of different types of neuro-fuzzy predictors are described and their performances are illustrated based on both the identification results from predictors and the measurements from an operational furnace.