Tailoring electric field signals of nonequilibrium discharges by the deep learning method and physical corrections

Abstract

Smart modulation of discharges is necessary to generate specific reactive species in an energy-efficient way. A physics corrected plasma + deep learning framework, the DeePlaskin, is proposed. This framework can be used for the nonequilibrium plasma systems that can be described by a global chemistry model (assuming global uniformity, e.g., in spark channels or the early afterglow of the fast ionization wave discharges). Knowing the kinetics scheme and the predefined temporal evolution of

Modulating sparks in a pulse train for repetitive and energy efficient plasma generation

“…Three codes are used in a combined manner in this study: the ZDPlasKin code for plasma chemistry [45], PASSKEy code for plasma transport [46,47], and DeePlasKin framework for inverse design [44].…”

“…A physics-corrected plasma + deep learning framework, DeePlaskin, was developed and tested for non-equilibrium plasma systems [44]. The framework proposes a 'predictor -corrector' approach combining the DNN and a global model to reconstruct the temporal profile of the reduced electric field E/N (and thus the potential drop on the electrodes) based on pre-defined temporal evolution of target species.…”

Review: Machine learning for advancing low-temperature plasma modeling and simulation

“…In this section, advances in ML for chemical engineering of plasma systems are explored. Zhu et al 89 developed a DL framework called DeePlasKin for non-equilibrium plasma systems that can be described by a global chemistry model. The model is based on a predictorcorrector approach detailed in Sec.…”

“…As highlighted in Sec. 2.2, Zhu et al 89 investigated a global chemistry model (with the possibility to interface with a Boltzmann solver) for a DBD as a function of the electric field. A discrete time-stepping predictor/corrector scheme has been suggested: (a) a reconstruction step to obtain a guess of the next reduced electric field E∕N value for a selected target density (at the next time step).…”

“…As highlighted in Sec. 2.2, Zhu et al 89 . investigated a global chemistry model (with the possibility to interface with a Boltzmann solver) for a DBD as a function of the electric field.…”