Impedance Matching Optimization Based On MATLAB

Open Access
Li, Kaiming
Graduate Program:
Electrical Engineering
Master of Science
Document Type:
Master Thesis
Date of Defense:
April 12, 2013
Committee Members:
  • James Kenneth Breakall, Thesis Advisor
  • Impedance matching
  • Global Optimization Toolbox
  • bandwidth
In the design of RF radiators, impedance matching has been costing a significant amount of time and energy. Approaches to solve impedance matching problems mainly rely on computer aided numerical optimizations nowadays. An approach based on the MATLAB Global Optimization ToolboxTM is proposed in this thesis. This approach combines brute-force techniques and the Real Frequency Technique, which are the two main components of modern impedance matching methods. The proposed approach allows the designer to choose a desired circuit topology, and use actual measurement impedance data from a candidate 80-meter high-frequency (HF) dipole antenna to get the optimum matching network. After comparing the performance of different algorithms at a single frequency, the simulated annealing algorithm is adopted as the solver in the Global Optimization ToolboxTM. Four combinations of inductor and capacitor (L only; C only; LC in series and LC in parallel) are filled in topology blank boxes, and then the optimizer compares all possibilities and presents the best matching network result. The Global Optimization ToolboxTM from MATLAB is the main software tool used in this thesis. The results are presented respectively for an L network, a T network, a Pi network, and 5-element and 6-element network topologies. The bandwidth can be increased from 3.4% to at least 14.4% after the optimization, with a VSWR of 2:1.