Ruihao Zhang / Beihang University (Beijing University of Aeronautics and Astronautics)
Meilin Wen / Beihang University
Modern electronic devices demand ever-smaller, higher-performance printed circuit boards (PCBs), yet miniaturization and complex service environments exacerbate failure risks. We first identify key failure modes under real-world conditions, build life-prediction models via multivariate nonlinear regression, and quantify reliability through uncertainty analysis. Next, we set optimization objectives and constraints based on failure-mode frequency, severity, and coupling, then formulate the coupled multi-objective model. An enhanced NSGA-II—using Lagrange relaxation to decouple constraints—expands the search space and improves solution diversity. Validation on a model-aircraft drive-control PCB demonstrates the method's effectiveness in balancing performance and reliability.