A C-Band Relativistic Backward Wave Oscillator Optimized by Multiobjective Algorithms

The parameters of high-power microwave sources exhibit high sensitivity and discontinuity regarding their power efficiency, operation frequency, and other key performances. Currently, particle in cell (PIC) simulation in conjunction with intelligent optimization algorithms is frequently utilized to optimize device parameters but there is still the problem that fast power ramp-up and frequency control cannot be achieved at the same time. In pursuit of an effective resolution, this article undertakes a comparison of three major classes of evolutionary algorithms within the domain of artificial intelligence during the design process of a C-band relativistic backward wave oscillator (RBWO). Through a new design method of segmented optimization strategy using the OMOPSO and RSEA algorithms, a novel RBWO structure has been attained. This structure attains an output power of 2.3 GW, functions at a goal microwave frequency of 4.3 GHz, and achieves an energy conversion efficiency of 46% in simulation under a magnetic field of 0.28 T. This novel optimization strategy is characterized by rapid optimization speed and favorable optimization efficacy.