Skip to Main Content
Considering factors such as economic costs and lives, an unreliable transportation system is more likely to cause severe consequences. Therefore, reliability optimization of transportation systems has attracted much attention over the past several decades. The traditional reliability optimization design is usually focused on redundancy allocation or reliability redundancy allocation. In practice, the operation process usually has a significant influence on the transportation system lifetime. By combining linear programming (LP) and extreme learning machine (ELM), a two-stage approach is proposed to optimize the transportation system lifetime, in which a semi-Markov model (SMM) is used to model the operation process. In the proposed method, we first formulate the optimization problem as an LP model, and the LP algorithm is utilized to search for the approximate optimal state probabilities. After data production and sample selection, ELM is trained with the produced training data and used to predict the optimal sojourn time distribution parameters. Applications on three different cases demonstrate that a higher lifetime can be ensured for the transportation system by using the proposed method.