New passivity criteria for uncertain neural networks with time-varying delay

This paper deals with the problem of passivity analysis for neural network with both time-varying delays and norm-bounded parameter uncertainties. A remarkable approach is proposed for constructing a novel Lyapunov–Krasovskii function involving triple integral terms. It does not requiring all the symmetric matrices to be positive definite. Due to the triple-integral terms and relaxation on the positive-definiteness of every Lyapunov-matrix, the conservatism of the results can be successfully reduced. Finally, numerical examples are given to demonstrate the effectiveness of proposed techniques.