Study On Thf Methods Of Cost Stimation For Government Invested Projects In Case Of Small Amounta Of Data

With the number of government investment project increasing, and the increasing of investment ceaselessly, the constructions have come out a series of questions, but also because of its uniqueness by the widespread concern. For example, the question in the control of investment of project of "three super" universal existence; and the operation and maintenance cost of the completion project which have put into using is far more than the initial investment. Investigated the reason, it is mainly because of the present stage of our country project investment estimated by the whole process of investment estimation methods, namely to construction oriented mode. It is easy to produce two phenomena: first, each party in engineering project investment control is in a fragmented state; second, the initial construction investment is often minimized which lead the operation maintenance cost to rise considerably. In each stage of the project, investment estimation phase of the project total investment have a high influence degree by more than80%, therefore, an accurate determination of the construction project investment estimation is very important.In this study based on whole life cycle cost of investment estimation method which has been more mature development in a foreign country was introduced to our country government investment project, take full account of project costs incurred in its various stages of whole life cycle from the project early stage until the end-of-life dismantling stage, and strive to make it minimum. In the present study, without affecting the precision of the investment estimation, the whole life cycle cost estimation model is simplified by significant theory, and the government investment project is divided into three cases to study. When the project has a large number of similar finished project, using back propagation neural network method to calculate the proposed project significant cost items (CSIs) investment and and significant factor (csf); when the project has none similar completed project cases, the fuzzy inference system in which the fuzzy language form expert experience quantification and participate in operation is used to estimate the project of CSIs and csf, calculate the total life cycle cost, and the two methods were evaluated separately.In the research, the government investment projects which have small amounts of similar completed project were analyzed in detail, which is the focus of this study. When the project exists little similar completed projects, use the fuzzy clustering method on the proposed project and similar completed project to extract CSIs, using the completed engineering data to calculate csf, through the use of the whole life cycle investment estimation model to predict the project investment estimation, and Verify the applicability of the fuzzy clustering method in this case by the actual case.Finally, it is pointed out that the possible contribution and achievements and development needs of the deficiencies of the study, hope that to find out the solutions for the actual construction of a series of problems of our country government investment project through this research, and provide a more scientific and reasonable investment estimation method, and provide the basis to make some contribution in their investment control.