Orthonormal basis function fuzzy systems for biological wastewater treatment processes modeling

In this paper, fuzzy models with orthonormal basis functions (OBF) framework are employed for modeling the nonlinear dynamics of biological treatment processes. These models are consisting of a linear part describing the system dynamics (Laguerre filters) followed by a non-linear static part (fuzzy system). The training procedure contains of two main steps: 1) obtaining the optimum time-scale and the order of truncated Laguerre network as the linear part and 2) defining membership functions, corresponding rules and adjusting the consequent parameters of fuzzy system as the nonlinear part. A comparison between the responses of the developed model and the original plant was performed in order to validate the accuracy of the developed model.