Research On Approaches Of Dynamic Estimation For Construction Project Costs Based On Bayesian Network

With the rapid development of the national economy, construction industry has entered a new era by leaps and bounds, a large number of domestic and foreign construction companies added to the market competition. Construction industry has low profit, fierce competition, and faces the complex market environment, enterprises are facing pressure to survive the enormous construction, so most construction companies will cost management as the core of enterprise management. The cost forecast is not only the starting point for construction enterprises cost management, the primary link for the project cost plan preparation and implementation of cost control, but also the main basis for construction companies bidding decisions, for improving the survival and development of enterprises of construction capacity and market competitiveness has very important significance. Cost forecast existing home and abroad following main issues:(1)Many construction enterprises have a weak awareness of projects cost.,the cost management system is not formed.(2) Cost forecast method behind and the low level of cost management.(3) One-sided deal with cost management, ignoring the unified antagonistic relationships among quality, duration, cost, safety and standardization construction in the cost forecast process.Scholars using variety of models and methods to predict and study the construction cost of the project, such as neural network, gray theory, earned value analysis, fuzzy theory, genetic algorithm, influencing factors, BOQ valuation analysis, multiple linear regression method, geological prediction methods, Bayesian networks and traditional risk assessment methods.This methods were good exploration to predicting the construction projects cost and achieved some success, but still flawed following aspects:(1) Previous studies have static prediction can only be achieved on the cost of construction projects, and dynamic forecast construction costs can not be achieved at any point in time; therefore, these costs apply prediction methods are limited to the construction tender stage, but not fully applied to guide the construction dynamic control project cost risk.(2)Research methods mostly did not consider the complex interaction relationships between the multiple factors in the formation progress of cost, and this complex relationships affect the mechanism of final construction project cost, therefore difficult to guide and improve cost control process.This article will ISM methods and Bayesian networks combined to analyze construction project cost factors, determine the impact of factor on the cost of the degree of influence and relationships, constitute ISM, ISM model based on Bayesian network analysis and forecasting, the cost of construction projects dynamic prediction model was constructed, and through project case validate the model. This study mainly has the following advantages and achievements:(1) Construction project cost forecasting method based on Bayesian network thinking clear, easy way, in practical work is also very easy to operate, the results reflected the degree of matching well with the actual situation, high reliability, but also in projects implementation process for immediate dynamic forecasting costs, so it has good prospects in the field of construction project cost management.(2)Through forward analysis of forecasting model for construction project cost based on Bayesian network can be quickly solve the expected cost of project, while the reverse analysis of Bayesian network model can be quickly calculated causing risk factors of project costs risk, On this basis can achieve dynamic early warning of project construction cost, and to provide effective guidance for projects cost management and cost control.