基于凸-改进遗传算法的圆阵列方向图联合优化

Joint Optimization about Pattern Synthesis of Circular Arrays Based on Convex Optimization and Modified Genetic Algorithm

针对圆形阵列方向图具有较高旁瓣的问题,提出一种新的基于凸优化和改进遗传算法的优化方法。该方法首先采用遗传算法将阵元位置和阵元权值作为优化变量,以最小化波束方向图峰值旁瓣为目标函数进行联合优化,既增加了变量的自由度,又符合理论意义上的全局寻优;同时,为了避免算法的早熟收敛,对基本遗传算法进行了必要的改进。然后采用凸优化方法对阵元权值进行二次优化可进一步降低旁瓣电平,与传统方法相比能够明显提高优化效果的稳定性。仿真数据证实了该方法的有效性和正确性。

To solve the high sidelobe level of circular arrays, an optimum method based on genetic algorithm and convex optimization is presented in this paper. First,it makes the location of the array element and the coefficient as joint variables and minimizing the highest sidelobe as its fitness function via genetic algorithm. And the traditional genetic algorithm must be modified in order to avoid premature convergence. And then,the object function can be farther minimum via convex optimization. Compared with the existing pattern synthesis method,it can enhance the stability of the optimum results. Computer simulation demonstrates the efficiency and accuracy of this method.