采用遗传算法优化局部放电超高频Hilbert分形天线

Application of Genetic Algorithms in Optimization of Partial Discharge Ultra-high Frequency Fractal Hilbert Antenna

为了获得性能最优的局部放电(PD)超高频(UHF)检测Hilbert分形天线,提出将遗传算法用于PD UHF Hilbert分形天线的优化设计。通过分析影响天线性能的结构参数,确定了其用于遗传算法的编码解码方式,设计了适应度函数,建立了UHF分形天线优化工程,并获得一组性能最优的天线参数;根据优化结果加工制作了原型天线。搭建PD试验平台,分别采用现有的分形天线及优化设计的4阶Hilbert分形天线检测PD信号。实测结果表明:当频率在300 MHz^1 GHz范围内时,天线的驻波比都<2,即在0~1 GHz的频率范围内天线具有700 MHz的检测带宽,达到了宽频带超高频天线的基本性能要求。与现有天线相比,优化后的天线具有更高的检测灵敏度和更宽的检测带宽,可有效应用于PD UHF信号的现场检测,获取更全面的PD信息。

In order to achieve Hilbert fractal antenna with best performance for partial discharge（PD） ultra-high- frequency（UHF） detection, we proposed to apply the genetic algorithms in the optimization and design of PD UHF Hil- bert fractal antenna. By analyzing structure configuration parameters that influence antenna＇s performance, we designed the encoding： and decoding method suitable for genetic algorithms and its fitness function. Then we established the opti- mization engineering for Hilbert fractal antenna and obtained a group of optimal performance parameters of Hilbert fractal antenna. On the basis of the optimized results, we fabricated prototype antenna. Actual PD experiments were car- ried out to detect PD signals using existing fractal antenna and optimized forth-order Hilbert fractal antenna respectively. The measurement results show that the voltage standing-wave ratio of antenna in the frequency of 300 MHz- 1 GHz is less than 2. When the frequency is less than 1 GHz, the detection bandwidth can reach 700 MHz, which meets the basic performance requirements of broadband UHF antenna. Compared with traditional antenna, the optimized Hilbert fractal antenna has higher detection sensitivity and wider bandwidth. Thus, the optimized antenna can be effectively applied to PD UHF filed detection and can obtain more comprehensive PD information.