Research on Prediction of Architectural Engineering Cost based on the Time Series Method

Abstract

Cost plays a very important role in architectural engineering, and the methods to predict the cost rapidly and accurately are needed. On the basis of the valuation mode of quantities bill, this paper proposed the time series method for predicting project cost. According ti the theory of test, modeling and prediction, the paper carried out time series analysis, taking steel price for example. It withdrew the steel price tendency item, then conducted modeling using the price residual item, finally established the model of steel price. In line with the model, the steel price of the first quarter in 2014 was forecast. By comparing with the real price, it shows that the error was in 5 percent, which proves the model is correct and the time series method based on the valuation mode of quantities bill can be adopted widely to predict manufacturing cost.

Determination of iron procurement strategy for manufacturing companies

“…the price of iron ore is dependent on the price of other commodities. Looking into the long-run dependence and causality between prices of crude oil and precious metals (gold, silver, platinum, palladium, steel, and titanium), Shafiullah et al, 2021 Since steel is also included as a significant cost of construction projects, it is important to forecast its price accurately; to do so, Zhang (2015) first conducted a time series analysis of steel prices and used the autoregressive moving average (ARMA) model for the future price, with the forecast results being considered accurate and reliable based on the tests conducted. To develop three models that use artificial neural networks for forecasting future prices of steel rebar in the context of the Egyptian construction industry 6 months ahead, Shiha et al ( 2020) used Microsoft Excel, NeuralTools software, and Python programming language in Spyder software to apply historical data on prices of steel and cement as well as macroeconomic indicators in Egypt from May 2008 to June 2018, where these proposed models can be potentially useful tools for forecasting and quantifying price fluctuations.…”