基于神经网络的成品物资采购价格研究

Research on Purchase Price of Finished Materials Based on Neural Network

能源行业数字化转型和当前国际原料价格波动双背景下,电力企业物资采购价格风险较大,为减轻企业经营风险,保障区域用电安全,电力企业通过构建多因素和基于时空特征提取的时间序列模型对国际铜价格进行预测,在此基础上基于回归模型构建原材料与成品招标价格的关系,从而强化对物资采购价格的把控和决策分析,实现对合理招标采购价格匡算、未来物资采购价格趋势预判以及招标价格区间设置,从而提升资源利用效率,保障物资供应质效。

Under the dual background of the digital transformation of the energy industry and the current international raw material price fluctuations, the price risk of power enterprises’ material pro-curement is relatively large. In order to reduce the business risks of enterprises and ensure the safety of regional electricity consumption, power enterprises have constructed multi-factor and spatio-temporal features based extraction methods. The time series model predicts the interna-tional copper price. On this basis, the relationship between the raw material and the finished product bidding price is constructed based on the regression model, so as to strengthen the control and decision-making analysis of the material purchase price, and realize the estimation of the reasonable bidding purchase price, trend prediction of future material purchase price and bidding price range setting, so as to improve resource utilization efficiency and ensure the quality and ef-ficiency of material supply.