Renewal Management Framework for Urban Rail Transit Assets

Abstract

Decision-making surrounding asset renewal is essential for the efficient use of renewal resources and safe operation of urban rail transit. In this study, major problems in the current management of urban rail industries in countries with the same problems as those in China were analysed, and in response, a renewal management framework based on service life estimation was proposed to provide adequate decision-making support for urban rail transit assets. In this framework, the cumulative failure frequency of an asset is used to indicate its health condition, and considering the uncertainties and heterogeneities in the deterioration process of assets, a Poisson–Weibull process model-based methodology was developed to estimate the service and residual lives of each asset, which are then employed in analysing its renewal demand and renewal period. Finally, the model is validated through an empirical study of rail renewal in the Beijing Metro. Our evaluation demonstrates that the proposed framework can estimate each asset’s service life accurately and can be used by asset management personnel to establish reasonable renewal plans and provide decision-making support for a scientifically informed resource allocation, thus mitigating major problems in current management practices.

The impact of summer heatwaves on railway track geometry maintenance

“…A Markov model follows the transition between different condition states, representing the degradation process in an exponential distribution 25 . Condition states are defined by discretised track geometry measurements, or other classes of track quality 26 . There are however two assumptions in a Markov model that limit their effectiveness in representing the track degradation process as it is currently understood; that transition probabilities are constant, and that future states depend only on the current state.…”

“…25 Condition states are defined by discretised track geometry measurements, or other classes of track quality. 26 There are however two assumptions in a Markov model that limit their effectiveness in representing the track degradation process as it is currently understood; that transition probabilities are constant, and that future states depend only on the current state. As the track degradation is observed to be dependent on the maintenance history (particularly tamping activities) the size of the Markov model grows dramatically to represent all possible transition rates.…”