Study On Cost Forecasting Of Overseas Project Based On Principal Component Regression Model

Over the past 30 years of reform and opening up,a large number of Chinese-funded enterprises have gone abroad,made use of foreign resources and markets,and carried out overseas investment work through direct foreign investment and contract engineering projects.They have made significant achievements in transportation construction and hydropower construction.In the face of the domestic industry with overcapacity and fierce competition,in order to realize the company's third leap-forward development,Z Company actively adjusted and transformed the industrial structure and accelerated the development of overseas projects.Relying on the company's resources and platforms,combined with the Mozambican resource market,Z company plans to invest in cement projects in Mozambique.For project investment,effective project cost control prediction can provide scientific basis for formulating reasonable and effective production plans and scientifically formulating bid decisions and investment decisions.Cost forecasting has become an important part of the project company's implementation of the project's scientific management costs and the preparation of cost plans,and it has been increasingly valued by managers.This paper aims at the current status of research on overseas engineering project cost forecasting,combined with the status quo of overseas engineering projects,and uses SPSS statistical analysis software to establish an overseas engineering project cost forecasting model based on principal component regression.First of all,this paper introduces the methods of cost forecasting,principal component analysis model,multiple regression model,and Z Company Mozambique A project.Then,according to the characteristics of overseas engineering projects,through the research and statistics of related documents and the consultation and summarization of relevant experts,the influencing factors of the selected overseas project cost are: project construction area,project construction period,national per capita GDP,and international steel The price index,theexchange rate of RMB against the U.S.dollar,the living standards of residents in the region,the level of outsourcers,the expectation coefficient of earnings,and the risk level coefficient,and sample data of 26 similar projects were selected for modeling.Due to the existence of multiple collinearity between the project cost impact factors,the regression model is difficult to solve such problems.Finally,this paper introduces a principal component regression model to predict costs.The principal component regression model combines the characteristics of principal component analysis and multiple regression models.By performing component extraction and performing multiple regression on the presence of co-linear indicators,the multicollinearity problem in modeling can be solved by such operations.Through the establishment of the principal component regression model for coefficient significant test,multi-collinearity diagnosis,fitness test and F test,we can see that the model passed various statistical tests,indicating that the established model is effective and feasible.Using the principal component regression model of overseas project cost forecast established in this paper to predict the Z Company Mozambique A project,the predicted value of the project cost is 16203.57 million US dollars.