Optimized design of a relativistic diffraction generator based on POD/PSO method

Abstract

When designing devices for high-frequency bands and high-power, it is inevitable to use overmoded structures. This paper proposes a proper orthogonal decomposition (POD) and a particle swarm optimization (PSO) collaborative algorithm based on particle-in-cell (PIC) simulation, which enables mode control and efficiency enhancement in overmoded high-power microwave sources. The algorithm is based on the orthogonality principle of electromagnetic eigenmodes, utilizing the field distributions of eigenmodes as basis functions to analyze the time-varying electromagnetic fields in three-dimensional PIC simulations. Furthermore, the objective function is designed to achieve co-optimization of both mode characteristics and efficiency performance. Combined with the PSO algorithm, an automatic optimization design is carried out for a Ka-band relativistic diffraction generator. The research results show that the POD/PSO collaborative optimization algorithm based on the particle-in-cell simulation method can converge to the global optimum and effectively improve the device efficiency. After optimization, the device efficiency (structure 1) with TM01 as the dominant mode is 26.7%; the device efficiency (structure 2) with TM11 as the dominant mode is 21.5%; the device efficiency (structure 3) under the joint dominance of TM03 and TM02 modes is 37.7%; and the device efficiency (structure 4) under the collaborative operation of TM02, TM03, and TM01 modes can be increased to 41%. Finally, compared with the independent PSO algorithm without mode analysis capability, the POD/PSO algorithm has better global search ability and converges to the optimal solution.