一种基于混合优化算法的阵列波束形成方法  

A new mixed optimization algorithm for array pattern synthesis

作　　者：王英民[1] 郁彦利[1] 齐华[1] 

机构地区：[1]西北工业大学航海学院,陕西西安710072

出　　处：《哈尔滨工程大学学报》2010年第7期832-836,共5页Journal of Harbin Engineering University

摘　　要：提出一种波束图设计的数值方法.该方法是组合运用了自适应阵原理和遗传算法的混合二阶优化算法.方法第一步是基于自适应理论,假定阵列为一自适应阵,通过在其波束图的旁瓣区布置大量的干扰信号以实现旁瓣控制.通过多次调整干扰强度,可以获得初始波束图.根据初始波束图建立新的期望波束图,基于估计波束图和期望波束图之间差别,建立误差能量函数.采用标准遗传算法通过最小化误差能量函数来实现波束图优化设计.双圈圆环阵的设计实例说明了算法的有效性.设计波束图的具体参数表明了其优良的性能.对于任一给定阵,该方法可设计给定指向方位、具有更低旁瓣的波束图.如果不能设计出期望特性的波束图,该方法可设计出最佳可得的波束图.A numerical pattern synthesis algorithm was formulated based on an adaptive notched noise（ANN） and a genetic optimization procedure.For a given array,this technique allows one to attain a set of weight coefficients that shape the array with a designed main beam in a given direction and lower sidelobes.The algorithm may search for attainable optimization patterns for the specific elements if the desired pattern cannot be found.The new algorithm as formulated has combined adaptive array principles with a genetic technique and a mixed two-step optimization approach.The first step was based on adaptive theory and the elements of the given array were assumed to be the elements of an adaptive array.The desired direction of the main beam was chosen as the steering direction and sidelobes were controlled by assigning a number of the interfering noises throughout the so-called sidelobe area.An initial weight coefficient or pattern was obtained after several iterations.Then an energy function was derived,based on the differences between desired pattern and the initial pattern.A simple genetic algorithm was used to minimize the energy function.The effectiveness of the algorithm was demonstrated on a sample double-circular array.

关 键 词：混合优化算法 阵列波束形成 遗传算法 

分 类 号：TB566[交通运输工程—水声工程]