Research On Target Cost Estimation Of Construction And Installation Of Real Estate Residential Projects

With the financing difficulties of real estate enterprises,cash flow tension and profit rate decline,cost control has become one of the core tasks of real estate enterprises.At present,the main investment products of China’s real estate enterprises are still residential projects.The main cost management method is target cost management.Target cost estimation is the key step of target cost management,and its accuracy can ensure the effective implementation of target cost management.Most real estate enterprises take the target cost estimated in the preliminary design stage as the baseline of cost control,and the construction and security cost estimation in the target cost estimation is a major part of inaccurate estimation with heavy workload and many details to be considered.Therefore,this paper studies the construction and security target cost estimation in the preliminary design stage of real estate residential projects.First of all,through the relevant literature research and to the real estate enterprises in the form of questionnaire survey and interview,analysis of the current causes of the preliminary design stage construction and security target cost estimation is not accurate two main reasons are : lack of previous residential project cost data for estimation reference;the current construction and security target cost estimation method has high requirements on the work experience of personnel cost estimation.Secondly,this paper summarizes the estimation methods of the separation of the construction and security cost subjects and the construction and security target cost,and analyzes that the estimation of the target cost of construction and decoration projects is most dependent on the cost work experience of the estimated personnel.Therefore,the residential construction and decoration projects are regarded as the construction objects of the SVM and BP neural network estimation models.On this basis,the specific content of the content index and the unilateral index of the construction and security cost database is studied in detail,which provides ideas and references for real estate enterprises to carry out the construction and security cost data precipitation of previous residential projects.Finally,the cost data of 35 residential projects built or under construction in Nanning are collected.Based on the research results of cost database,the ABC classification method and literature survey method are used to extract 16 quantitative indexes that affect the cost of residential building decoration engineering.As the characteristic indexes of SVM and BP neural network estimation models,the target cost estimation models of residential building decoration engineering based on SVM and BP neural network are constructed respectively.The 35 residential cost data are used as the sample data of the model,and the SVM and BP neural network estimation models are trained and simulated in MATLAB.By comparing and analyzing the simulation test results of the two models,the model with good prediction effect is selected as the target cost estimation model of residential building decoration engineering.After the estimation model is determined,different characteristic indexes are selected for training and simulation test in MATLAB to further study the influence of different characteristic indexes on the prediction effect of the estimation model.