The Cost Risk Of Construction Project Forecasting Based On PSO-SVM

Construction Project is a systems engineering with dynamic character, the project in the implementation process is affected by many factors, it exists many uncertainties,which has great risks at the same time.Comparing with developed countries, there is less risk awareness about risk management of construction project; also entire risk management expertise is far from the level in these countries. In such as case, this paper is focused on the forecast of the risk of various costs at the construction stage from the viewpoint of contractor, which touches both of the theory and practice under the basis of summarizing and analyzing the findings in this research area by the experts in China and outside China.First of all, this paper introduced the definition of cost risk at the phase of construction. Then it placed emphasis on meticulous and completely identification of cost risk at the phase of construction through Systematic Analysis.Project materials’ study and Survey at the spot during the construction, considering the actual situation of construction projects. Cost risk factors of construction project are identified from the owner, constructer, supervision unit, environment, cost objective and management six factors. On this basis, we built the cost risk assessment index system of the construction. Superior performance of support vector machine in predicting system is used to deal with the uncertainty risk factor in the construction projects, and then combines with Particle Swarm Optimization to build the Particle Swarm Optimization-support vector machine model. In order to prove the feasibility and effectiveness of the model, the predicted result is compared with support vector machine parameters. Finally, the thesis made a training test and an applied analysis of the specific sampled data to testify the built prediction model.