研究目的
To identify and analyze design trade-offs for miniaturized CP antenna including the antenna capability for maintaining small size while retaining acceptable levels of other performance figures.
研究成果
Multiobjective optimization is a useful tool for assessment of antenna capabilities with respect to a set of conflicting objectives. The Pareto front provides comprehensive data concerning capabilities of a given antenna structure, which cannot be achieved using single-objective approaches. Despite usefulness of numerical optimization for reduction of antenna dimensions, selection of appropriate topology is crucial for size reduction.
研究不足
The actual reflection and AR bandwidths are wider than the frequency range of interest because the objective formulation refers to the maximum levels, not specific thresholds for |S11| and AR. The change of the structure size across the Pareto front is lower compared to other structures reported in the literature.
1：Experimental Design and Method Selection:
Multiobjective optimization is used to identify and analyze design trade-offs for miniaturized CP antenna. A population-based metaheuristic algorithm is employed to obtain a set of designs representing the best attainable compromise between the imposed requirements. The algorithm is executed on a cheap approximation model and the results are further corrected to bring them to the EM model accuracy level.
2：Sample Selection and Data Sources:
A planar CP antenna is considered for the analysis. The antenna high-fidelity EM model Rf and its low-fidelity counterpart Rcd are implemented in CST Microwave Studio.
3：List of Experimental Equipment and Materials:
CST Microwave Studio is used for EM modeling and simulation.
4：Experimental Procedures and Operational Workflow:
The design process involves reducing the initial design space, sampling the search space and acquiring Rcd data, constructing a kriging interpolation model Rs, finding initial Pareto set through MOEA optimization of Rs, and refining selected designs.
5：Data Analysis Methods:
The analysis is performed for a 15-parameter structure with elliptical ground plane slots. The computational cost corresponds to around 449 Rf model evaluations.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
