Generation of meter-scale nanosecond pulsed DBD and the intelligent evaluation based on multi-dimensional feature parameter extraction

This study introduces a novel meter-scale dielectric barrier discharge (m-DBD) reactor designed to generate large-scale, low-temperature nanosecond pulsed discharge plasma. By employing a modularized gas path, this reactor enables a comprehensive analysis of discharge patterns and uniformity using multi-dimensional discharge parameters. Simulation results reveal optimal gas distribution with ten gas holes in the variable plate and a 40 mm slit depth in the main reactor. Besides, a diagnosis method based on electro-acoustic-spectrum-image (E-A-S-I) parameters is developed to evaluate nanosecond pulsed m-DBD discharge states. It is found that the discharge states are closely related to the consistency of segmental discharge currents, the fluctuation of acoustic signals and the distribution of active particles. Machine learning methods are established to realize the diagnosis of m-DBD discharge pattern and uniformity by E-A-S-I parameters, where the optimized BPNN has a best recognition accuracy of 97.5%. Furthermore, leveraging nanosecond pulse power in Ar/m-DBD enables stable 1120 × 70 mm² discharge, uniformly enhancing hydrophobicity of large-scale materials from a 67° to 122° water contact angle with maximal fluctuations below 7%. The modularized m-DBD reactor and its intelligent analysis based on multi-dimensional parameter provide a crucial foundation for advancing large-scale nanosecond pulsed plasma and their industrial applications.