Rules Mining Intelligent Algorithm Applied In Optimization Of Frequency Selective Surface Absorbing Structures

Frequency selective surface(FSS) is a kind of two-dimensional periodic array structure, periodically arranged by a particular cell pattern. Since FSS characterizes selective permeability for electromagnetic wave, it is playing an increasingly important role in the military, life, and production. In this paper, the "parameters–performance" rules mining module for FSS and the intelligent optimization algorithms with good performance were investigated and designed, depending on the plug-in of microwave Intelligent Integration Platform(MIDIS) extension features.Based on MIDIS platform, we designed and completed the rules mining module and intelligent optimization algorithm module in plug-way. Firstly, using the For algorithm module of MIDIS Platform, can calculate batch quantities and store results having identical format in a database. Many FSS characteristics, the position, depth, bandwidth of the absorption peaks and variable parametric regularity, can be advantageously analyzed by using the rules analysis module. The designed A-EDA algorithm, EDA algorithm, and GA-EDA algorithm, can optimize FSS absorptive performances efficiently, establishing a "rules mining- Algorithm optimization" optimization framework.With optimization framework reasonably utilized, single square ring FSS absorber properties can be optimized and acquire a good broadband absorbing effect(-10 dB from 4.0 to 8.0 GHz). Then, the rules analysis module can contribute to analyze the regularities of effects on absorptive bandwidth which arise from the differences between outer and inner diameter or the space between unit and outer diameter, etc. The final optimized t absorptive structure through A-EDA algorithm, EDA algorithm, and GA algorithm has been fabricated and measured. The tested results were-10 dB in a range of 4.4-8.9 GHz. As well, the three optimization algorithms were compared and it was shown that A-EDA algorithm has the excellent ability to optimize the performance of designed FSS.With optimization framework reasonably utilized, double square ring FSS absorber properties can be optimized and the reflectivity goal is to be-10 dB from 3.0 to 10.0 GHz. Then, the regularities of effects on absorptive bandwidth, which arise from the width of outer and inner rings, or the space between unit and outer diameter, etc, can be obtained with the rules mining module. We used the A-EDA algorithm, the EDA algorithm, and the GA algorithm to optimize double square ring FSS; measurement results were-10 dB in a range of 2.6-10.2 GHz. Through the comparison of varied frameworks, the high efficiency of the framework, "rules mining- Algorithm optimization" was demonstrated.