Experimental realization of physical reservoir computing-based mechano-intelligence in self-adaptive phononic metastructures

Abstract

This research experimentally investigates the integration of mechano-intelligence into mechanical metastructures for self-adaptive wave control. We created a phononic metastructure prototype utilizing periodic buckled beam modules that has highly adjustable wave propagation characteristics via length reconfiguration using a linear displacement actuator. By utilizing the physical reservoir computing framework, we show that the proposed metastructure can recognize and self-adapt to different inputs by making decisions on appropriate actuations to reconfigure itself to achieve an intelligent wave blocking task. Overall, this research provided a promising approach for constructing and integrating functional mechano-intelligence in structures harnessing physical computing and learning, and created a new direction for the next generation of adaptive structures and material systems.