基于改进BP神经网络的药品集团采购供应链需求预测研究

Research on Demand Forecasting of Pharmaceutical Group Purchasing Supply Chain Based on Improved BP Neural Network

摘要: 药品集团采购模式的高质量发展需要解决药品集团采购供应链末端医疗机构的需求预测问题。针对医疗机构当前药品需求预测不准确影响报量的问题，探讨了如何利用改进BP神经网络提高药品预测的准确性。通过对药品需求量的影响因素进行相关性分析和主成分分析，选择了最重要的影响因素作为输入变量，并运用“二分分割法”确定最佳隐含层节点数，构建了改进BP神经网络需求预测模型，并以某医疗机构销售的药品阿奇霉素为例进行了仿真测试与数据验证。结果表明，改进的BP神经网络模型在预测药品需求量上具有较高准确性和稳定性，其误差保持在合理范围内。该研究不仅提升了药品需求预测的精确度，还为医疗机构制定科学的采购计划和库存策略提供了重要依据。

Abstract: The high-quality development of pharmaceutical group procurement mode needs to solve the demand prediction problem of medical institutions at the end of pharmaceutical group procurement supply chain. Aiming at the problem that inaccurate drug demand forecast in medical institutions affects the reported quantity, this paper probes into how to improve the accuracy of drug forecast by using improved BP neural network. Through correlation analysis and principal component analysis of the influencing factors of drug demand, the most important influencing factors are selected as input variables, and the optimal number of hidden layer nodes is determined by “bisection method”, and an improved BP neural network demand forecasting model is constructed. Taking azithromycin sold by a medical institution as an example, the simulation test and data verification are carried out. The results show that the improved BP neural network model has high accuracy and stability in forecasting drug demand, and its error remains within a reasonable range. This study not only improves the accuracy of drug demand forecasting, but also provides an important basis for medical institutions to formulate scientific procurement plans and inventory strategies.