论文部分内容阅读
分析了简单趋势、时间序列、神经网络等预测方法，结合组合预测和滚动优化的思想提出了适合于某钢厂冷轧发货预报的预测策略，给出了严格的基于实际生产数据的仿真研究．仿真结果表明，依据模型提出的预测策略是可行的
The prediction methods such as simple trend, time series and neural network are analyzed. Combined with the idea of combined forecasting and rolling optimization, a forecasting strategy suitable for cold rolling delivery forecast of a steel mill is proposed. The simulation research based on actual production data is given . The simulation results show that the forecasting strategy based on the model is feasible