On-line linear system parameter estimation using the neo-fuzzy-neuron algorithm

A method for estimating the parameters of a single input single output (SISO) model is proposed and discussed. The new method is based on the neo-fuzzy-neuron algorithm and has the property of fast convergence, which makes it suitable for online estimation. In order to evaluate the estimator effectiveness, it was applied to obtain the parameters of a second-order filter. Simulation and experimental results are presented.