| Evolutionary algorithm can improve the efficiency and performance of antennadesign, so it was applied to the antenna design, this approach was the developmenttrend for modern antenna to design and optimize, and had important theoreticalsignificance and application prospect.The theory of microstrip antenna and reconfigurable antenna were described inthe thesis. Based on genetic algorithm、 HFSS-MATLAB-API and HFSS anoptimization scheme was put forward. Optimization steps of the scheme weredisplayed together with the definition of gene sequences, individual fitness value,design of fitness function and consideration of genetic strategies.Upon the scheme a set of optimization program had made out. After optimized thebandwidth of window antenna was about10%and E shaped antenna was about34.5%,which realized the design requirements of ultra-wideband. A frequency reconfigurablemicrostrip antenna was obtained, which can work dynamically at Lband、S band or Xband, radiaton pattern cover on the upper half space. The PIFA antenna which hassingle frequency was optimized with portable dimensions at the working frequency. Onaccount of single frequency that double frequency was optimized sucessfully, tworesonant frequencies were1.79GHz and2.44GHz, which approach the target1.8GHzand2.4GHz accuratly. Through these examples the feasibility and effectiveness of thescheme was proved.Finally, an ultra-wideband sector microstrip antenna was optimized by the neuralnetwork. The neural network and genetic algorithm was compared and the respectiveadvantages and disadvantages was analysed. |
PDF Full Text Request