Probabilistic design optimization of shipboard radar mast by adopting RBFN meta-model and various reliability methods

This study presents a probabilistic design optimization method for enhancing the design safety of shipboard radar mast, which accommodates navigational communication equipment such as radar scanners and antennas. Such structure requires not only robust vibration and strength performance but also minimized weight to reduce marine pollution and increase operational efficiency. Given the lack of definitive classification rules for radar mast structural design, this study employs various reliability analysis methods. A radial basis function neural-network (RBFN) meta-model, generated from Design of Experiments data, was utilized for optimization and reliability analyzes. The probabilistic design optimization problem was formulated to determine the random design variables such that the weight is minimized subject to the probabilistic constraints of vibration and structural strength performance. Various reliability analysis methods such as adaptive importance sampling, first-order reliability method, mean value first-order second moment method, and second-order reliability method were compared to identify the best approach for the probabilistic design optimization. The study concludes by identifying the reliable probabilistic optimal method for improving design safety relative to deterministic design optimization results.