基于PCA与Elman网络的机械产品质量建模

多源多工序是现代制造过程的一个显著特征,针对该特征提出了一种基于主成分分析,和Elman网络的机械产品质量建模的方法。通过对样本数据空间的主成分分析,能够保证在信息损失最少的情况下,对高维变量空间进行降维处理,减少样本数据间的相关性。应用典型的动态回归Elman神经网络,实现复杂非线性系统进行建模和预测;还将其应用于冷轧带肋钢筋的机械性能预测中。

Quality Modeling of Mechanical Product Based on PCA and Elman Neural Network

Multi-source and Multi-working procedure are one of the notable feature of the modern manufacturing operation. A way is proposed for quality modeling of mechanical product that based on PCA and Elman network. To analysis the sample data space by PCA can assume that it can lower the dimension of high variant space and eliminate the relativity of sample data. Elman network is a typical dynamic regressive neural network. It can model and predict the complicated non-linear system. This way is applied to predicting mechanical properties of cold-rolled ribbed steel wires and bars which get good effect.