Antenna placement on complex platforms by multi-objective genetic algorithms

Two new AP-GA techniques that minimize calls to CEM solvers have been developed. They extract as much information as possible from a single CEM solver run, to serve in the assessment of multiple design candidates, instead of just one. The proposed schemes leverage multilevel (inner-outer) Pareto GAs, along with approximate objective function evaluations to reduce the cost of forward CEM analysis without sacrificing the convergent nature of the GA. Even though the proposed AP-GAs apply to a wide variety of scenarios, here, they are elucidated in the context of placing/synthesizing loaded monopole constellations on an aircraft.