DeePulse: An in-situ ML-assisted optimization framework of plasma sources for hydrodynamics control andenergy conversion

Abstract

Smart control of the driven power is one of the key issues of plasma source optimization. We present here, from code to device, a closed-loop optimization framework, to achieve in-situ plasma source control by dynamically adjusting the driven voltage parameters. The framework has been used in SDBD for hydrodynamics control in a turbulent flow wind tunnel and in DBD for chemical active species production.