Machine Learning Assisted Array Synthesis Under Mutual Coupling and Platform Effects

Abstract

An efficient machine learning assisted array synthesis (MLAAS) method is proposed for practical array designs under mutual coupling and platform effects. By introducing machine learning methods, the impact of the electromagnetic environment on the antenna element is learned and utilized to build surrogate models for active element patterns and S- parameters. Compared with conventional methods, the proposed MLAAS is able to achieve great prediction accuracy based on limited numbers of full-wave simulations. Moreover, the algorithm is able to deal with array design problems with variable element numbers, which is validated using a practical antenna array design task.