摘要
给出了一种基于实数编码遗传算法的天线方向图综合方法。在遗传算法模型中采用了向下、向上外推和非一致杂交算子等交叉技术 ,并结合内插、交换等多种技术形成综合交叉方式。该算法克服了已有算法早熟、对初始群体依赖性等缺陷的同时 ,较大幅度地提高了算法的收敛速度和可靠性 ;通过在目标函数中加入零陷方差项克服了现有算法零陷不均衡的缺陷。计算机仿真结果表明 ,与现有算法相比 ,该方法用于天线方向图综合具有收敛速度快、零陷均衡、可靠等优势。
An array antenna pattern synthesis based on a real-valued genetic algorithm is presented. The genetic algorithm model adopts some cross methods such as downwards extrapolate, upwards extrapolate and nonuniform cross operator. Moreover, this method integrates with interpolation and exchange to form compositive integrated cross method, by which it can get rid of the prematurity and the dependence of the convergence of conventional genetic algorithms on the initial population. At the same time, variance of nulls is joined the objective function to overcome the imbalance of nulls. Computer simulation proves that this method has advantages of fast convergence, nulls balance and reliability in array antenna synthesis compared existing algorithms.
出处
《电波科学学报》
EI
CSCD
2004年第2期182-186,共5页
Chinese Journal of Radio Science
基金
教育部资助优秀青年教师计划资助 (编号 :2 0 0 0 16 36 )
关键词
遗传算法
天线方向图
无线通信
智能天线
实数编码
方向图综合
Antenna arrays
Computer simulation
Genetic algorithms
Optimization
Signal encoding
Wireless telecommunication systems
