Energy-efficient adaptive control using spiking neural networks and biologically inspired learning mechanisms.
AIAS LabPurdue University Northwest
This research investigates spiking neural networks and neuromorphic learning for adaptive control and estimation. Event-driven computation is combined with control-theoretic analysis to develop learning architectures that are computationally efficient while retaining stability and safety properties.
Topics include spiking actor-critic control, adaptive internal synaptic weights, event-driven observers, human intent estimation, and safety-critical neural control for autonomous and robotic systems.