Railway vehicle robust design optimization via a polynomial chaos approach

This study addresses the critical challenge of ensuring railway vehicle system (RVS) safety and comfort when traveling on short-radius curved tracks. Traditional deterministic optimization methods often fail to account for the inherent uncertainties in design variables (DVs), which can significantly affect a vehicle’s performance. The objective of this work is to develop a robust optimization framework that simultaneously considers RVS safety and comfort while accounting for these uncertainties. To achieve this, an algorithm combining a multi-objective improved colonial competitive approach (MICCA) and the polynomial chaos expansion method (PCEM) is developed. This robust optimization process integrates RVS safety and comfort, their standard deviations, and the uncertainties of the DVs into a single framework. The obtained results demonstrate that the robust design significantly reduces the sensitivity of the railway vehicle’s performance to design variable uncertainties, outperforming deterministic approaches and existing literature findings.