The goal of this paper is to introduce some identification and control systems concepts to the field of high power microwave (HPM) tubes. These concepts are well known to the control systems community, but have not yet been fully exploited within the HPM community. The simpler mathematical approach used is contrasted with the more physical modeling, using first principles and advocated by experimental results. The paper also reports on a preliminary application of these ideas to the Sinus-6 electron beam accelerator. We present simulation results which show that a simple nonlinear model using static neural networks is sufficient to accurately model the input/output behavior of the Sinus-6 driven backward wave oscillator (BWO).
展开▼
