Research On The Model And Example Of Construction Cost Estimation In Feasibility Study Stage Based On Principal Component Regression

In recent years,China's construction industry is developing rapidly.At the same time,construction enterprises are also facing fierce competition.In order to save cost and improve efficiency,construction enterprises must manage project cost scientifically.The estimation of project cost is the first link of cost management,which lays the foundation for the follow-up cost management.At present,there are some problems in the method of engineering cost estimation,such as low estimation accuracy and complex application conditions,which are not conducive to the management of engineering cost.In view of this situation,after studying a large number of documents,this paper applies the principal component regression method to the field of cost estimation.First of all,through the Internet cost information platform to collect project cost cases and analyze the influencing factors,13 characteristic indexes are initially selected and quantified,and 5 typical characteristic cost indexes with significant influence are selected through step-by-step preliminary regression.Then principal component analysis is used to deal with the multicollinearity of the regression model,and the principal component regression model is established and restored.Finally,after the stability test,the regression model is applied to a practical engineering case to prove its effectiveness.The research results of this paper show that:1.in the feasibility study stage,a small number of typical characteristic cost indexes can be extracted from dozens of comprehensive and systematic cost factors.In this paper,five typical characteristic cost indexes such as main structure,foundation type,decoration level,installation engineering cost and non competitive cost can be extracted.2.Through these typical characteristic cost indicators,the principal component analysis and multiple regression methods are used to build the principal component regression estimation model,which has good effect and good fitting degree in the feasibility study stage.3.The estimation accuracy of the model is more than 90%.Three randomly selected engineering cases show that the average errors are 2.4%,14%and 7.6%respectively.The accuracy of the model is much higher than the standard error rate of 20%,which proves that the model can be applied to the engineering cost estimation practice.The principal component regression estimation model applies mathematical statistics theory to cost estimation,with higher precision and easy to master in practice.The principal component regression estimation model provides a new idea for engineering estimation practice,and has certain reference significance.