流程型企业主生产计划优化的ARMA-BP模型

为了解决流程型企业主生产计划(MPS)时段中计划参数与实际参数的不一致性问题,提出了自回归滑动平均模型(ARMA)与反向传播(BP)人工神经网络(ANN)的集成优化模型。基于主生产计划时段长度与产量之间的映射关系,利用平均时段长度折合产量法(OCM-ATS),该模型可用于分别逼近和预测主生产计划时段的长度时序和产量时序。给出的例子表明,该模型预测主生产计划时段的参数(计划参数)与实际参数的相对误差不超过3%。

ARMA-BP model for optimizing master production schedule in process enterprises

To solve the inconsistency between planning parameters and practical parameters under a time horizon of master production schedule(MPS) in process enterprises, an optimized model of integrating auto regressive moving average(ARMA) model with back propagation(BP) algorithm of artificial neural network(ANN) is proposed.Based on the mapping from the span to the output of MPS time horizon,the ARMA-BP model can respectively approach and forecast the span and the output time series by virtue of output conversion method with average time span(OCM-ATS). An example indicates that relative errors between forecasting parameters(planning parameters) and practical parameters are all less than 3% by applying ARMABP model.