Research on A New Method to Estimate Whole Life Railway Project Cost Based on CS, FC and FIS theories

摘要

Accurate estimation of preliminary cost in feasible study stage is an important precondition for cost control of whole project. The present cost estimation methods all estimate it through setting up linear functions of variables according to the data and information of similar performed projects, or through the quota made by the quota and standard department of Ministry of Railways. When there are no historic data and information, preliminary cost is often made by adding up the experts' estimation. However, those result in big errors and unstable reliabilities, which is one of the reasons that cost of many performed railway projects are out of control. Based on the WLC and CS theory, a set of methods are set up in this paper. When a scheduled railway project has no similar performed projects, we select the CSIs of it based on the CS theory, and divide CSIs into two parts, that is, the ones which have similar items with the performed projects and the others which have no similar items. The former costs could be forecasted by the FC (Fuzzy Clustering), and the latter could be estimated through Fuzzy Inference System, which fully takes advantage of experts' experience. This method not only simplifies the workload and procedure but also enhance the accuracy of cost estimation.