Whole Life Cycle Cost Management

The subway projects have huge investment,long operation cycle and high operation and maintenance cost,and the high cost will restrict the development of subway projects.Therefore,Therefore,the design and planning of the project from the perspective of whole life cycle cost management is an urgent problem for China's subway projects.The whole life cycle cost management is the research of decision making stage,design stage,implementation stage,completion acceptance stage,operation maintenance stage and scrap stage.The goal of modern construction projects is to make decision by holding the concept of sustainable development;to minimize the whole life cycle cost,and maximize the value of construction projects;to optimize cost management during the operation and maintenance period,to ensure the recycling of resources and minimal environmental impact.Therefore,we need to analyze the cost problem from the perspective of the whole life cycle of the project.A big problem to face is huge data processing and heavy calculation of the whole life cycle cost.Therefore,the focus of the paper is to mine the most effective information from the massive data,and estimate the cost through grasping of these significant information.So as to reduce the computational difficulty and improve the computational efficiency.First of all,this paper studies the correlation theory of the whole life cycle cost,including the phase division of the whole life cycle cost,the composition of the whole life cycle cost and the mathematical model of the whole life cycle cost.From the perspective of the formation time,it is clear that the whole life cycle cost is composed of the initial construction cost and the future operation and maintenance cost.The whole life cycle cost mathematical model is used as the basis to calculate the construction cost and operation cost respectively.For the initial construction cost,the similar engineering with similar engineering characteristics and similar cost was excavated by using the rough set software Rosetta.Using the " mean value theory" method to obtain the significant items of various similar projects,to verify the applicability of the significant cost theory.Then,the civil cost is estimated and summarized into construction costs with other costs.For the futureoperation maintenance cost,this paper will use operation data and the engineering characteristics affecting the operation cost of the completed project as the research object,and build BP neural network.Using MATLAB software,data are trained and tested to obtain the operating cost.Finally,the initial construction cost and future operation maintenance cost are summarized,and the whole life cycle cost is obtained.And the method is applied to the subway project,through the example of Beijing subway line 4,the estimation method used in this paper can estimate the cost of the whole life cycle of the project,greatly reduce the workload,improve the efficiency,and the accuracy is also allowed.Through the study on the bill of quantities of 15 subway lines,it is found that the items with significant impact on the cost account for 18.82 % ~ 28.35 % of the total cost,which accords with the thought of the significant theory.The cost estimated through the significant items is also within the allowable error range of the investment estimate.To build the BP neural network by using the historical operation cost data and the corresponding influence factors,the error of the simulated operating cost and actual historical operating cost data is also within acceptable range.The study found that CSIS model and BP neural network model can mine and grasp small part of significant data from the mass data,Using a more convenient method to reduce the workload of the project cost estimation,can also ensure that the whole life cycle cost estimation value of the project is with in the acceptance range of investment estimation error.At the same time,through the verification of the subway project,this model can also be used in highway,bridge and subway projects,but the methods of implementation still need to be analyzed.