Nonlinear System Identification of a Lower Limb Model by Fuzzy Wavelet Neural Networks

This paper presents the identification of a simulated nonlinear system that represents the dynamical mechanism of the human lower limb. The study and application of this model may have a relevant importance in the research area of rehabilitation of patients suffering from any kind of paralysis of their lower limbs. Here, a Fuzzy Wavelet Neural Network (FWNN) is used to identify the lower limb model under study. In order to evaluate the FWNN model, it was validated in two distinct situations. Firstly it was considered that the original model does not suffer any modification in its parameters and, in the second case, the viscous coefficient was reduced. In this way, it was possible to analyze the FWNN model robustness in terms of this parameter change. The performance of the FWNN was also compared with other two neural network structures: Multilayer Perceptron (MLP) and Wavelet Neural Network (WNN).