An Efficient Optimization Method for Slow Wave Structures Based on Surrogate Model for TWT Applications

Slow wave structures (SWSs) are the most important component in traveling wave tubes (TWTs). High interaction impedance and low dispersion of the SWS are essential for achieving high output power and wide operating bandwidth in TWTs. For now, the optimization of the SWS still highly relies on electromagnetic (EM) simulations, posing challenges in terms of time and effort. To address this issue, an optimization framework for SWS using the surrogate model combined with the differential evolution (DE) algorithm and the penalty function (PF) is developed in this article. Support vector regression (SVR) is introduced as the surrogate model to replace the time-consuming simulations with good accuracy. By combing the surrogate model with DE and PF, the proposed method is able to optimize the interaction impedance with specific phase velocity with low dispersion. For verification, the proposed method is applied to the circular helix SWS and the T-shaped staggered double-vane (T-SDV) SWS as examples. The results clearly show the effectiveness and advantages of the proposed method.