Passivity analysis of memristive neural networks with different memductance functions

• A new memristive system with different memductance functions is formulated.
• Some novel control schemes are proposed for state-dependent memristive system.
• The unified form of criterion is an effective methodology for memristive system.
• The method for qualitative analysis of memristive system is efficacious.

Memristive neural networks have captured the attention of physicists, biologists, ecologists, economists and social scientists. In this paper, we formulate and investigate a class of memristive neural networks with two different types of memductance functions. Some succinct criteria in terms of linear matrix inequalities for the passivity are proposed. Meanwhile, based on the derived criteria, some stability criterion are obtained for the memristive neural networks. These theoretical analysis can characterize the fundamental electrical properties of memristive systems and provide convenience for applications.