Improvement of atmospheric jet-array plasma uniformity assisted by artificial neural networks

Atmospheric pressure plasma jet (APPJ) arrays have shown a potential in a wide range of applications ranging from material processing to biomedicine. In these applications, targets with complex three-dimensional structures often easily affect plasma uniformity. However, the uniformity is usually crucially important in application areas such as biomedicine, etc. In this work, the flow and electric field collaborative modulations are used to improve the uniformity of the plasma downstream. Taking a two-dimensional sloped metallic substrate with a 10° inclined angle as an example, the influences of both flow and electric field on the electron and typical active species distributions downstream are studied based on a multi-field coupling model. The electric and flow fields modulations are first separately applied to test the influence. Results show that the electric field modulation has an obvious improvement on the uniformity of plasma while the flow field modulation effect is limited. Based on such outputs, a collaborative modulation of both fields is then applied, and shows a much better effect on the uniformity. To make further advances, a basic strategy of uniformity improvement is thus acquired. To achieve the goal, an artificial neural network method with reasonable accuracy is then used to predict the correlation between plasma processing parameters and downstream uniformity properties for further improvement of the plasma uniformity. An optional scheme taking advantage of the flexibility of APPJ arrays is then developed for practical demands.