Research On The Cost Estimation Model Of The Utility Tunnel Based On Support Vector Machine

With the development of economy and society,the construction of utility tunnel has been carried out in all major cities.Utility tunnel construction develops rapidly,construction method is also tending to variety,and constructing utility tunnel engineering needs long time to and high investment cost,utility tunnel belongs to the underground engineering,total construction investment is affected by multiple factors,such as construction environment and price level,all of these put forward higher request for utility tunnel engineering cost management.At present,the estimation method of utility tunnel is too general or too time-consuming to meet the requirements of actual project cost management for accuracy and efficiency.Therefore,it is necessary to build a new fast and accurate cost estimation model combined with utility tunnel engineering construction,the project cost composition,management present situation,based on the recent existing engineering data with the reasonable mathematical modeling methods.Based on the characteristics and cost characteristics of the utility tunnel,this study proposes a fast and reliable estimation model of the new utility tunnel project based on the cost data of the built project.First,in view of a small number of utility tunnel have been built,cost characteristics of many influence factors,which are frequently used in engineering cost estimation model,on the basis of selection can effectively deal with small sample,nonlinear and high dimensional data of the support vector machine(SVM)as a modeling basis.Secondly,according to the construction cost structure and the characteristics of the project,this paper analyzes and extracts the characteristic indexes of the cost estimation of the comprehensive pipe gallery project.On the basis of the above research,constructs the utility tunnel construction cost estimation based on support vector machine(SVM)model,and use the principal component analysis(PCA),genetic algorithm and the application of basic model for two improvements.On the one hand,the main component analysis method is used to carry out the data preprocessing process,and the original data information can be maintained to the maximum extent,and the comprehensive and independent new indexes are obtained.On the other hand is for support vector machine(SVM)model parameters selection depending on the experience,the question of subjectivity,genetic algorithm is proposed for parameter optimization of the model improvement method,nearly step to improve the stability of the model and estimation precision.Finally,the above research is integrated and extended to construct the new composite model,and the data of the construction project of the utility tunnel are simulated and verified.The simulation results show that compared with the traditional model,the new model has higher estimation precision and more stable estimation effect,which can meet the requirements of the construction cost management of the utility tunnel.The model can solve the problem of cost management in the investment decision-making stage of the utility tunnel,which has great practical significance to cost control and engineering risk reduction.