A Numerical Mixed Optimization Algorithm for Array Pattern Synthesis

摘要

The problem of array pattern synthesis with desired shapes has received much attention over years. A numerical pattern synthesis algorithm is presented. The new algorithm combines adaptive array principle with Genetic technique and a mixed two-step optimization approach is formulated. The first step is based on adaptive theory and the elements of the given array are supposed as the elements of an adaptive array. The sidelobes are controlled by assigning a number of interfering signal throughout the sidelobe region. An initial weight coefficients or pattern is obtained after several iterations and a new desired pattern is constructed. Then an error energy function (objective function) based on the difference between estimated pattern and the desired pattern is introduced. A simple Genetic algorithm is used to optimize the pattern by minimizing the error energy function. The effectiveness of the algorithm is demonstrated on an example of a doublecircular array. Design results are given in details and show very good performance. For a given array, this technique allows one to attain a set of weight coefficients that shape the array with designed main beam in a given directions and lower sidelobes. The algorithm may search for optimization attainable patterns for the specific elements if the desired pattern cannot be found.