Reliable and Robust Optimization of the Planetary Gear Train Using Particle Swarm Optimization and Monte Carlo Simulation

Uncertainties in design represent a considerable industrial stake. Controlling the reliability and robustness of a mechanical system at the level of design has become necessary in order to control these uncertainties. Using the theory of probabilistic design optimization, the present work reports on the application of the concept of reliability-based robustness on minimizing the weight of a planetary gear train (PGT). The optimum combination of reliability and robustness for the minimum weight of the PGT was found using an optimization algorithm based on Particle Swarm Optimization (PSO) and Monte Carlo Simulation (MCS). The algorithm was developed by combining the propagation of uncertainties with the optimization of the function objective within a single probabilistic model. The results show that a reliability-based robust design offers a better alternative to the traditional deterministic design models. An equivalence between the proposed method and the deterministic method is established and approved.