State estimation for complex networks with randomly varying nonlinearities and sensor failures

Abstract

The current study is focused on the l2−l∞ state estimator design for the discrete-time complex networks with sensor failures and randomly varying nonlinearities. Bernoulli process is adopted to describe the randomly varying nonlinearities, and the norm-bounded uncertain model is used to deal with the sensor failures. Then, a set of sufficient conditions are provided to guarantee that the estimation error system is stochastically stable with the prescribed l2−l∞ property. Then, using the linear matrix inequality method, the estimator gains are obtained. Finally, the effectiveness of the proposed new design method is illustrated through a numerical example. © 2016 Wiley Periodicals, Inc. Complexity 21: 507–517, 2016