Protocol-based fault detection for discrete-time memristive neural networks with quantization effect

Abstract

This paper is concerned with the event-triggered fault detection filter design problem for discrete-time memristive neural networks with measurement quantization. Aiming at saving communication bandwidth and improving communication efficiency, a new component-based adjusted dynamic event-triggered protocol is proposed. The transmission status of each measurement component can be determined based on the individual prescribed triggering condition. The quantizer with variable density is introduced to meet different performance requirements, in which the quantization densities vary among a finite set of modes. By constructing an augmented Lyapunov functional, an asynchronous filter is designed to achieve the desired performance while assuring the stochastic stability of the filtering error system. Finally, a simulation example is provided to verify the effectiveness of obtained theoretical results.