Abstract
► We formulate a multiobjective robust design optimization of rail vehicle systems. ► The deterministic solutions can be altered by the design parameters uncertainty. ► An original method that combines Monte Carlo and Genetic Algorithms is developed. ► A robust optimization gives solutions less sensitive to the design uncertainty.
This work deals with the multiobjective robust design optimization of rail vehicle systems moving in short radius curved tracks. Two criteria are considered simultaneously, i.e., safety (considered by the derailment risk) and comfort given by noise level. The authors show that the deterministic optimal solutions, for the nominal design parameters, can be altered seriously by the design parameters uncertainty. The authors of this paper propose an original algorithm that combines Genetic Algorithms and Monte Carlo Simulation in order to be used for the robust multiobjective optimization of the rail vehicle design. The obtained solutions, presented by design vectors of the rail vehicle, are analyzed in terms of performances and robustness. The authors show that the robust multiobjective optimization can yield solutions less sensitive to the design parameters uncertainties.