| 摘 要: 针对传统电磁优化方法耗时长、效率低的问题,提出一种基于粒子群优化算法(PSO)与耦合参数模型相结合的微波滤波器优化方法。该方法利用耦合参数模型实现结构参数与频率响应的映射关系,在信赖域内通过PSO算法对结构参数进行搜索优化。通过与基于耦合参数模型的深度强化学习(DQN)方法进行对比实验,该方法在优化性能方面表现出色,优化时间较DQN方法减少32%。实验结果表明,该方法能够实现快速可靠的微波滤波器设计,具有较高的实际应用价值。 |
| 关键词: 电磁(EM)建模 神经网络 耦合矩阵 粒子群算法PSO |
|
中图分类号: TP391
文献标识码: A
|
| 基金项目: 国家自然科学基金项目(32360437);甘肃省高等学校产业支撑计划项目(2021CYZC-57) |
|
| Optimization Method for Microwave Filters Based on Particle Swarm Algorithm and Coupled Parameter Model |
|
LIU Linglong, LIU Wenyuan, YAN Bo
|
(School of Electronic Information and Artificial Intelligence, Shaanxi University of Science and Technology, Xi’an 710021, China)
liu_linglong2025@163.com; liuwenyuan@sust.edu.cn; ryanbo118@163.com
|
| Abstract: To address the issues of time-consuming and inefficient traditional electromagnetic optimization methods, this paper proposes a microwave filter optimization approach that integrates the Particle Swarm Optimization (PSO) algorithm with a coupled parameter model. This method establishes a mapping relationship between structural parameters and frequency responses through the coupled parameter model, and employs the PSO algorithm to search and optimize structural parameters within a trust region. Comparative experiments with the Deep Q-Network (DQN) method based on coupled parameter models demonstrate that this approach exhibits superior optimization performance,reducing optimization time by 32% compared to the DQN method. Experimental results indicate that this method enables rapid and reliable microwave filter design with significant practical application value. |
| Keywords: electromagnetic modeling neural network coupling matrix particle swarm optimization(PSO) |
