Keywords: railway renewal, railway maintenance, optimization, genetic algorithm.

Most railroads are used very frequently and they play an important role in transporting hundreds of people and large amount of goods on a daily basis. The integrity of railroad tracks is challenged by the friction forces caused by trains passing over them and the corrosive effects of the surrounding environment. As with all other equipment, railroad tracks require regular maintenance in order to function properly. There are several maintenance operations that aim to increase the lifespan of a railroad, however in some cases, neither of these operations becomes sufficient and the entire railroad section needs to be renewed. Railways organizations use different proportions of ordinary maintenance and periodic renewal with little consensus as to the best combination. The cost-effectiveness of emphasizing one method over the other had not been analysed using empirical data. It is important to define the correct time for track maintenance and renewal works. Track maintenance and renewal works carried out at the correct time is crucial to realize an efficient and optimized maintenance and renewal work plans and thus increase the life of the track components. Whereas carrying out maintenance and renewal works too late is certainly unsafe and the railway track get older consequently the maintenance and renewal costs increases exponentially.

In this paper, a genetic algorithm based decision support system is designed for parameter analysis to decide the correct maintenance and renewal method for the optimization of railway track maintenance and renewal activities. Genetic algorithms, which are strong adaptive optimization methods based on biological principles, were used to determine the optimized railway track maintenance and renewal works. Genetic algorithms were successfully applied in this study and reasonable results were found. The proposed method is based on replacing operations that are too costly (i.e. consume too many resources) with ones that are less costly but similar to the original operations. Before running the genetic algorithm, the user defines some constraints on the maximum amount of resources for the maintenance. While the optimization algorithm is running, it chooses solutions which are within the specified constraints, but also as close as possible to the ideal solution. Results of this study showed that with an optimal choice of population size and time, a best solution with the given constraints could be achieved as to the means of maintenance and the renewal of the tracks. With the increase of the sensitivity of the reference data, more accurate solutions can be achieved.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £110 +P&P)