Array synthesis under practical constraints is a vital design task. Traditional array synthesis methods usually deal with isolated antenna elements without considering mutual coupling (MC) or mounting-platform effects, which results in unacceptable degradation in practical array designs. An efficient machine learning-assisted array synthesis (MLAAS) method is introduced using efficient active base element modeling (ABEM). This method greatly extends the boundaries of practical antenna array synthesis from the perspectives of both accuracy and design freedom. Using much fewer samples than those in conventional MLAAS methods, all possible element designs are accurately modeled into one active base element (ABE). Compared with conventional active element pattern (AEP)-based methods, the ABEM aims to predict AEPs for elements with arbitrary allocations and electromagnetic (EM) surroundings, therefore offering more degrees of freedom for practical array designs. Four array design examples are used to verify the effectiveness of the proposed method.

An efficient multilayer machine learning-assisted optimization (ML-MLAO)-based robust design method is proposed for antenna and array applications. Machine learning methods are introduced into multiple layers of the robust design process, including worst-case analysis (WCA), maximum input tolerance hypervolume (MITH) searching, and robust optimization, considerably accelerating the whole robust design process. First, based on a surrogate model mapping between the design parameters and performance, WCA is performed using a genetic algorithm to ensure reliability. MITH searching is then carried out using a double-layer MLAO (DL-MLAO) framework to find the MITH of the given design point. Next, based on the training set obtained using DL-MLAO, correlations between the design parameters and the MITH are learned. The robust design is carried out using surrogate models for both the performance and the MITH, and these models are updated online following the ML-MLAO scheme. Furthermore, two examples, including an array synthesis problem and an antenna design problem, are used to verify the proposed ML-MLAO method. Finally, the numerical results and computation time are discussed to demonstrate the effectiveness of the proposed method.