Author(s)

J. F. Shen, Y. D. Xu & H. F. Li

Abstract

To make the track comprehensive maintenance plan reasonably is an important means to ensure the safety of train operation. The track maintenance plan model analysis shows that making the annual maintenance plan is a process of finding the global optimal solution in a high dimension space. Reducing the model dimension is the key to simplify the complexity of a model under the premise of keeping the integrity of the model, so the maintenance plan model is divided into two parts, the track unit selection model and MTT job assignment model. In this paper, the first model will be introduced in detail. In this model, three unit selection constraints are considered, the optimized object is each lot of units and the objective function is to get the maximized sum of the maintenance, improving quantity of all the lots that are included in the selected unit. These selected units being optimized will be used as the input parameters of the MTT job assignment model, and this can achieve the purpose of reducing the dimension of the latter model. In order to solve and verify the first model, a genetic algorithm is introduced, constraint conditions in the generation phase of the initial population are pre-processed. The results, based on data from Shanghai-Kunming railway line, show that the solving efficiency is several times higher than the enumeration method under the circumstance of ensuring the average error is less than 15%. This proves that the model is practical and the genetic algorithm is effective in efficiency and precision, so the model can be used as a rapid and efficient approach for making a track maintenance plan. Keywords: railway track, track longitudinal irregularity, comprehensive maintenance plan, unit selection, genetic algorithm.

Keywords

railway track, track longitudinal irregularity, comprehensive maintenance plan, unit selection, genetic algorithm.