A modified version of the multi-network neural model for an accurate nonlinear system modeling

This paper proposes a new idea for modeling complex systems: it is a modified version of the Multi-Network Neural Model ("MNNM") to solve the problem of nonlinear systems modeling. In fact, the Multi-Network Neural Model was carried out using an algorithm consisting on training simultaneously all local neural networks, then, the interpolation of these local networks via an appropriate fuzzy interference system. In this paper we try to improve the characteristics of the "MNNM" by training some fuzzy sets parameters at the same time with neural networks parameters. The experimental results show the effectiveness of this study compared to the "MNNM" classic one. This work will be supported by a second order non linear system.