Dynamic Adaptability and Prediction Optimization in Power Material Supply Chains using XGBoost

Accurate material demand forecasts are crucial for the effective preparation of power material supply plans. Collaborating with suppliers in the material supply chain and establishing a cooperation mechanism based on mutual trust, information sharing and risk exchange can synchronize operations and enhance management efficiency. Therefore, it is particularly important to thoroughly investigate methods for improving the dynamic adaptability of modern (intelligent) material supply chains and enhancing the management of enterprise material flow. Under this background, we have made research and got the following conclusions: (1) there are six stages in the life cycle of electric power materials. The cost percentage in the planning stage accounts for 5%; the cost percentage in the design stage accounts for 10%; the cost percentage of the manufacturing stage accounts for 37.5%; the cost percentage of the installation stage accounts for 20% and the cost percentage in the operation stage accounts for 15%. Different models predict power materials and XGboost has a better prediction ability. (2) Power material supply chain data quality management is mainly divided into material support system. The first-level index is 0.413, the second-level index system is 0.578 and the weight is 0.7096. The first-level index of logistics timeliness is 0.379, the second-level index is 0.469 and the weight is 0.2577.