The Research Of Construction Cost Estimation Based On Artificial Neural Network

All works of life have faced severe challenge since China entered WTO. Construction trade, which is one of the mainstay industries in our country, bears great shock under international competition. The management of construction cost is an important field in project management, and the management of project cost under a gradual reform, former simple ration management model translates into a new model that includes guideline of"unified quantity","supervised price","competitive cost". The new model gives attention to the situation of China. The construction estimation is a basic respect in construction cost management. The construction estimation not only provides basis for project authorization, but also is an important function of architectural design and is an important basis of cost control during the whole course.It's important to make an exact evaluation in the prophase of project, especially enter into the 21st century. On one hand, the competition during all real estate corporations is becoming more and more keen, the cycle of project orientation, decision-making, development and sell becomes more and more short. One the other hand, the construction cost evaluation as its function bases on drawings, although the evaluation method based on oration or real quantity can provide exact date to support decision-making, it spends long time, and it's reserve to the rule that"Time is life". Therefore, it becomes urgent to seek an effective method of construction evaluation at present. Artificial neural network is a popular method that solve complex problem in academia, especially it is good at calculating on non-linearity problem. This essay tries to use artificial neural network to evaluate construction cost.This paper firstly discusses the problems of construction cost management base on reading lot of literature, summarizes some abroad advanced cost mode, then discusses the orientation of cost management. In the context of former, the conception, extension and influent aspect of construction cost evaluation are expounded. Then some methods of construction cost evaluation at present are mainly introduced, its keystone, applied ways are expatiated, and its apply value in true-life is estimated, some cases are used to prove it.Artificial neural network is applied later, and its conception, basic keystone, common arithmetic rules are introduced, BP network mode and RBF network are applied, and some improved ways are used to enhance the quality of network, the comparative effect is pointed out. At last, all methods are debated and compared, the feasibility of ANN application in construction evaluation is proved, and then the deficiency and the further research of this paper are given out.