论文部分内容阅读
无缝钢管轧制过程机理复杂,部分状态参数难以在线检测,轧制过程具有多变量、强耦合、非线性等特点,导致其机理建模精度较低。为了提高轧制力模型的精度,提出一种结合机理计算与神经网络预测的轧制力建模方法。首先依据轧制工艺知识和经验分析轧制过程机理,建立轧制力机理模型;然后依据实际生产数据,通过灰色关联分析确定影响轧制力的主要因素;最后采用BP神经网络建立轧制力偏差预测模型,对轧制力机理模型计算结果进行补偿。仿真实验表明,该模型预测精度较高,可以满足工业现场的实际需求。
The mechanism of seamless steel tube rolling process is complex, some of the state parameters are difficult to detect on-line, and the rolling process has the characteristics of multivariable, strong coupling and nonlinearity, resulting in low accuracy of mechanism modeling. In order to improve the accuracy of the rolling force model, a rolling force modeling method based on mechanism calculation and neural network prediction is proposed. Firstly, according to the knowledge and experience of rolling technology, the mechanism of rolling process is analyzed to establish the rolling force mechanism model. Then, the main factors affecting the rolling force are determined by gray relational analysis according to the actual production data. Finally, the BP neural network is used to establish the rolling force deviation Prediction model to compensate the rolling force mechanism model calculation results. Simulation results show that the model has higher prediction accuracy and can meet the actual needs of industrial field.