摘要
提出了一种基于本征激励方向图综合的改进实数编码遗传算法。在遗传算法选择算子中引进分裂选择机制,较好地开发了解空间;交叉算子采用了基因交换、中间重组、算术交叉形成综合交叉方式,对解的开发深度明显优于单一的传统算术交叉;对变异概率采用渐变控制,有效地扩大了搜索空间,克服了传统算法早熟的现象,良好的搜索结果证明了算法的精确性。算法基于本征方向图综合思想,由于实测本征方向图包涵了互耦信息,有效地解决了阵列天线互耦问题,避免了阵列互耦校正的复杂过程,具有较大的实用价值。
A modified real genetic algorithm based on eigen-driven analysis method for pattern synthesis is presented in this paper. In the selection algorithm, the splitting selection mechanism is introduced to exploit the solution area better. A mixed crossover algorithm containing gene crossover, middle recombination and arithmetic crossover enlarges the searching area. Besides, it overcomes the prematurity of the convergence of conventional genetic algorithm. Excellent computer result proves the high precision of the algorithm. Because mutual coupling information is included in eigen-driven patterns, the algorithm based on eigen-driven patterns which are measured artificially in this paper solves the mutual coupling effects. So, this algorithm avoids complicating calibration process and has high practicality.
出处
《中国电子科学研究院学报》
2008年第3期321-326,共6页
Journal of China Academy of Electronics and Information Technology
关键词
互耦
副瓣
遗传算法
阵列天线
本征方向图
mutual coupling
sidelobe level
genetic algorithm
antenna array
eigen-driven patterns
