Research On Fuzzy Math And BP Neural Network In Highway Cost Prediction Model

Highway construction is closely related to economic development, in order to drive economic growth, the state has been committed to the construction of highway engineering, The highway is a top priority.Accurately estimate the highway project cost,make a rational and impartial evaluation of highway investment, not only effectively to control the highway construction,realize the maximization of the use of investment funds and avoid the waste of funds,But also to ensure the pre unit before the audit can have a general understanding of the total project cost, and is conducive to the smooth development of audit work.In addition, with the development of bidding system in our country,the current project budgeting method has low efficiency, long cycle and other issues,to a certain extent, need a new cost prediction method can accurate quotation and improve the competitiveness.In this thesis, according to the present situation of highway project cost forecasting, starting from the perspective of the construction unit, through to the highway engineering cost prediction at home and abroad related research,as well as in-depth analysis of the affecting factors of highway engineering cost,on the basis of the traditional prediction methods, establishes a prediction model based on fuzzy mathematics and BP neural network, realize the rapid prediction of highway engineering cost.Using relevant knowledge of fuzzy mathematics, choose those with similar to predict engineering project as the training sample, to realize the nonlinear mapping between the characteristic of highway engineering and construction cost by using the BP neural network,and then the case analysis, proved the practicability of the model.