The Study of Traveling Wave Tube Large Signal Model Based on Machine Learning

Abstract

Driven by success in areas such as computer vision and natural language processing, attempts have been made in this work to combine deep learning with the large signal of the Traveling Wave Tube (TWT) to assist in predicting the output performance of the tube and advancing the pre-design of TWT. By feeding the trained artificial neural network with several feature parameters, the output power with the tube length can be predicted.