Research On Forecasting Method And Influencing Factors Of Construction Steel Price

As the country continues to increase its investment in infrastructure and the rapid development of the real estate industry,steel consumption in the construction industry has shown a trend of increasing year by year.Rebar is one of the most important manufactured products of the entire steel industry and the main structural component of construction projects,accounting for more than half of the steel consumption in the construction industry.In recent years,the price of rebar is subject to changes in market supply and demand and related economic and trade policies.It is in a state of continuous fluctuation,which has brought greater price risks to construction steel companies.Therefore,a more important issue for the construction industry is how to avoid the price crisis caused by frequent changes in rebar prices.Through analysis of the existing research,it is found that most scholars are studying the influencing factors of the entire steel market price,and the application of the combined forecasting model is relatively few when conducting price forecasting research.Therefore,based on the existing research basis,this paper sorts out the key influencing factors of rebar price changes and introduces a combined forecasting model to study and predict its price,in order to provide reference basis and information support for investment control of construction steel enterprises and reduce investment risks..First of all,this paper analyzes the research trends in relevant fields at home and abroad,using literature research methods and combining the characteristics of the rebar market,mainly from the macroeconomics,market supply and demand,production cost composition and other factors,selected 22 influencing factors and In-depth analysis was carried out separately;secondly,relevant data were obtained from the National Bureau of Statistics,Chongqing Construction Cost Information Network and other websites,and correlation analysis was conducted to select 15 influential factors with strong correlation and statistical significance,and then gray correlation analysis was adopted The method further measures and analyzes the correlation between the 15 influencing factor indicators and the price of rebar,and digs out the key factors that affect the price of rebar to lay the foundation for the research on the price prediction of rebar;again,using statistical analysis and quantitative analysis,through the The volatility characteristics of the selected rebar sample data were studied,and the selection criteria of the prediction model were constructed,and three models of exponential smoothing model,ARIMA model and multiple linear regression model were selected to construct a single-item rebar price prediction model library.At the same time,in order to improve the accuracy of the model prediction results,the BP neural network and the principal component analysis method are introduced to establish a combined prediction model to form a rebar price combination prediction model library.Finally,empirical research is carried out on various prediction models,using the model accuracy evaluation index Judge the validity of each model.The study found that the three most important rebar price influencing factors are coke price,rebar export volume and manufacturing purchasing manager index.Among the three single-item prediction models,the prediction accuracy of the ARIMA model is the best,the prediction accuracy of the exponential smoothing model is slightly inferior,and the prediction accuracy of the multiple linear regression model is relatively poor.The prediction accuracy of the two combined prediction models is good,but the prediction accuracy of the combined prediction model of BP neural network is significantly better than the combined prediction model of principal component analysis.Taken together,the prediction effect of the two combined prediction models is significantly better than that of various single-item prediction models.Among the five prediction models,the prediction effect of the combined prediction model of BP neural network is the best.This article systematically analyzes the influencing factors of rebar prices,clarifies the degree of influence of each influencing factor,constructs two rebar price prediction model libraries,and analyzes the fitting effects of each prediction model through empirical research,and finally obtains the most suitable Price prediction model.This has positive and practical significance for construction steel companies and rebar manufacturers to master the price trend of rebar and avoid price risks in advance.