State estimation for T-S fuzzy Hopfield neural networks via strict output passivation of the error system

This article presents a new design scheme for the state estimator for Takagi-Sugeno fuzzy delayed Hopfield neural networks that uses strict output passivation of the error system. Based on Lyapunov-Krasovskii functional, Jensens inequality, and linear matrix inequality (LMI) formulation, a new delay-dependent criterion is proposed such that makes the resulting estimation error system exponentially stable and passive from the input vector to the output error vector. The unknown gain matrix of the proposed state estimator can be obtained by solving the LMI, which can be facilitated using existing numerical packages. We verify the effectiveness of the proposed state estimation method through a numerical example.