Prediction and Analysis of Wuxi Stainless Steel Market Price Based on ARIMA-BP Neural Network Combination Model

Abstract

: Stainless steel is not only a high-strength and high-toughness structural material, but also an economical functional material. My country’s stainless steel industry occupies a pivotal position in the fields of construction, transportation, energy, and packaging. It is also particularly important to study its price trends and internal laws. Therefore, this paper selects the price of 201 series hot-rolled and cold-rolled stainless steel and 304-series hot-rolled and cold-rolled stainless steel in Wuxi City as the research objects, and uses the time series model and BP neural network model to model them respectively, and predicts the prices of future stainless steel .Then a parallel ARIMA-BP combined model is established, and the weights of the two models are distributed by the advantage matrix method to obtain new forecast data. Comparing and analyzing absolute error and relative error, it is concluded that the accuracy of combined prediction model is higher than that of basic prediction models such as time series and BP neural network. In addition, this paper analyzes the factors influencing the fluctuations in the price of stainless steel, and the significant degree of correlation between stainless steel price and various factors is tested by the Pearson coefficient, and the possible reasons for it are explained. Finally, the future development of stainless steel is prospected.