Improved conditions for global exponential stability of a general class of memristive neural networks

• A switched memristive network cluster is formulated.
• Proper feature is being taken of multi-terminal memristor state.
• The free-weighting parameters are used to reduce the conservativeness.
• The convergence rate can be estimated via theoretical results.

Modeling and related characterization of memristive neurodynamic systems becomes a critical pathway towards neuromorphic system designs. This paper presents a general class of memristive neural networks with time-varying delays. Some improved algebraic criteria for global exponential stability of memristive neural networks are obtained. The criteria improve some previous results and are easy to be verified with the physical parameters of system itself. The proposed framework for theoretical analysis of memristive neurodynamic systems may be useful in developing nanoscale memristor device as synapse in neuromorphic computing architectures.