At present,the global steel industry is undergoing great recession,which leads to the fact that many enterprises,including several large Chinese iron and steel enterprises,fall in an overall loss.In this case,how to enhance the competitiveness of enterprises has become the research frontier of the management science. Improving customer satisfaction and reducing inventory costs are two primary means to achieve the goal.In this paper,the BP neural network is improved by using the learning rate adaptive algorithm,and then the improved neural network is used to simulate the production data of the steel plant by using the Matlab toolbox simulation. Adopting the trained network forecasting model,the actual safety stock is predicted.Results show that the proposed method can accurately and efficiently predict the security inventory needed by the welded pipe plant, and provide decision support for reasonable inventory to improve the efficiency of inventory control.%提升企业竞争力是管理科学的研究前沿,提高顾客满意度和降低库存成本是2个重要手段,库存控制已成为提高企业竞争力的关键因素.通过运用学习率自适应算法改进BP神经网络,然后用改进的神经网络并运用Matlab工具箱仿真实现对钢铁厂生产数据的网络训练,利用训练好的网络预测模型对焊管厂的实际安全库存进行预测.仿真及实际运行结果表明:该方法能够准确高效地预测焊管厂所需安全库存,可以为合理的库存提供决策支持,有效提高库存控制的效率.
展开▼
