Study On Optimization Algorithm Of Microwave Cavity Filter Structural Parameters

With the development of communication technologies, frequency band between signals will become more and more narrower. In order to suppress various interference signal and reduce the signal attenuation, scholars of the world have done a lot of research work on the design of microwave filter. Since microwave cavity filter has the advantages of high power, low loss, high inhibition, compact structure, and stable performance, it has been widely used in the communication system in recent years. At the same time, with the rapid development of computer technology, the combination of high-frequency electromagnetic simulation software and filter design make the design precision greatly improved and the design cycle of microwave filters greatly shorten. But for complex filters with many parameters to design, it is difficult for electromagnetic simulation software to meet the requirements of design precision in due time and sometimes even difficult to convergence.In order to solve these problems, this paper introduces an improved neural network and extreme learning machine (ELM) algorithm into the design of cavity filter and completed the rapid modeling of HFSS (High Frequency Structure Simulator) in Matlab environment in order to improve the efficiency of the filter design and shorten the design cycle. Firstly, this paper studies the basic principles of band pass filters, coupling structure between resonator cavities and the design of coupling port. And then analyzes and studies the optimization design methods of microwave cavity band pass filter using neural network, proposed genetic neural network(GA-BP) and parallel quasi-Newton neural network algorithm(PQN-BP) are proposed. The two optimization algorithms respectively used in the optimization of structure parameter of microwave cavity filter. And then introduces and analyzes ELM, kernel function, and quantum particle swarm theory (QPSO). Kernel Extreme Learning Machine (KELM) is introduced to the field of optimization of microwave filter structure parameter for the first time. Finally, an optimization model of microwave cavity filter based on WKELM with quantum particle swarm optimization algorithm is established. Comparing and analyzing the three algorithms proposed, QPSO-WKELM algorithm has the best effect. Thus, a microwave cavity band pass filter is designed using QPSO-WKELM. The simulation results show that the algorithm can achieve fast and accurate microwave filter design, compared with conventional filter modeling techniques, its accuracy and speed have been greatly improved.