Simultaneous topology and size optimization of locomotive structure using multinary genetic algorithms

Abstract

Structural optimization of a typical locomotive carbody subjected to various operational loads is investigated using a simultaneous topology and size optimization approach. The proposed approach builds on a string representation of the structure in which the presence or absence of a structural element between specific nodes, as well as its cross-section, are represented by integers. Strings representing various structure topologies are then evolved according to a modified Genetic algorithm with improved genetic operators. The efficiency of the proposed model is verified through applying it to optimize the carbody structure of ER24PC locomotive. Operational load cases and performance criteria are adopted from the European standard EN12663. It is shown that the results based on the present optimal design have higher safety factors compared to the original design without a significant increase in weight. Besides, the computational cost of the optimization process is shown to be considerably less than that of the classical binary Genetic algorithm.