论文部分内容阅读
为有效预测智能制造模式下的不确定性需求,提出自回归移动平均模型ARIMA和改进BP神经网络的组合模型,对预测数据中包含线性规律的Lt以及非线性规律的ε_t进行模拟和分析,以解决预测有效性和精度问题.通过数据样本构建,对ARIMA模型结构进行辨识,确定p,d,q参数,并对模型进行诊断和检验;在此基础上进行需求数据一次预测;通过连接权值的修正降低BP神经网络学习误差,并对一次预测结果与原需求数据样本存在的误差进行二次预测.实例数据分析表明:组合模型的预测精度较ARIMA模型有显著提高,因此组合预测模型在预测效果上具有合理性和有效性.
In order to effectively predict the demand of uncertainty in the intelligent manufacturing mode, a combined model of ARMAA and improved BP neural network is proposed. Lt of linear prediction and ε_t of non-linear rule are simulated and analyzed. Solve the problem of prediction validity and accuracy.Through the data sample construction, the structure of ARIMA model is identified, the parameters of p, d, q are determined, and the model is diagnosed and tested; once the demand data is predicted, , The second-order prediction of the error between the first prediction result and the original demand data sample is carried out.Example data analysis shows that the prediction accuracy of the combined model is significantly higher than that of ARIMA model, The effect is reasonable and effective.