摘要
Polymer composite materials are widely used in industries, such as transportation and renewable energy, due to their lightweight nature, high strength, and outstanding durability. Ensuring their long-term reliability under fatigue conditions is critical for safety and performance, and this requires efficient accelerated life testing (ALT) methods. The ALT aims to provide precise lifetime predictions while minimizing costs. However, existing approaches often rely on locally optimal designs depending on accurate guesses of model parameters, which is unrealistic given the inherent uncertainty by their values. To address this issue, this paper introduces a standardized minimax optimal design method for fatigue testing of polymer-composite materials. This method addresses parameter uncertainty by incorporating a range of possible parameter values, thus providing protection against worst-case scenarios. The optimal design incorporates a hybrid optimization strategy, combining a particle swarm optimization (PSO) algorithm with additional techniques to overcome challenges posed by the nondifferentiable criterion and multilayer nested optimization problem. Numerical results demonstrate that these standardized minimax optimal designs outperform conventional locally optimal designs and Bayesian optimal designs, offering improved efficiency and reliability. This work provides a practical and robust framework for assessing the long-term performance of polymer-composite materials in real-world applications.