Finite-Time H-Infinity Filter Design for Discrete Jump Time-Delay Neural Networks

摘要

This paper studies the problem of finite-time H-infinity filter design of discrete Markovian jump neural networks with time-varying delays and norm-bounded disturbance. The definitions of stochastic finite-time boundedness and stochastic H-infinity finite-time boundedness are first given. Then, an H-infinity state estimator is designed for the extended neural networks to ensure stochastic finite-time boundedness of the error dynamics with the prescribed disturbance attenuation level in a given finite-time interval. Furthermore, sufficient criteria are presented for the solvability of the state estimation problems by applying the linear matrix inequality technique and Markovian system approach. Finally, an example is presented to show the validity of the design method.