基于LSTM-JITRVM的冷轧轧制力建模方法研究

Research on Modeling Method of Cold Rolling Force Based on LSTM-JITRVM

在带钢冷连轧生产过程中,轧制力预测准确度直接影响产品质量。为提高轧制力预测准确度,提出了基于LSTM-JITRVM(long short term memory-just in time relevance vector machine)的轧制力模型。首先,使用循环自编码网络对输入数据进行深层次特征提取,然后使用局部离群因子算法判断测试样本与其邻域点是否属于同一分布,针对不同的分布使用不同的自学习回归模型进行拟合。仿真结果表明,该模型预测准确度可控制在3%以内,能够实现轧制力的高准确度在线预测。

In the process of strip cold rolling,the accuracy rolling force directly affects the product quality.In order to further improve the prediction accuracy of rolling force,a rolling force model based on LSTM-JITRVM(long short term memory-just in time relevance vector machine)is proposed.Firstly,the auto-encoding network is used to extract the deep-seated features of the input data.Then the local outlier factor algorithm is used to judge whether the test sample and its neighborhood points belong to the same distribution,and different self-learning regression models are used to fit different distributions.The simulation results show that the prediction accuracy of the model can be controlled within 3%,and the rolling force can be predicted online with high precision.