ABSTRACT
The common features of the two fuzzy models delays chain needed to store the past model output to be used as antecedents as the system evolves in discrete-time steps. This chapter focuses on building mathematical models of environmental systems, and analyses their behaviors by using the models. No special assumption is made to the model structure, which may be non-linear both in the past inputs and outputs, and in the parameters. The chapter uses ANFIS to revisit the function approximation problem. The tank exercise is now revisited by generating an ANFIS model. The chapter concludes the tank exercise by comparing the performance of the first and second order models. It can be seen that the simpler first-order model produces a performance comparable to that of the second-order model, though having a much simpler structure that makes its use much more attractive from the computational viewpoint.
