论文部分内容阅读
张力效应和速度效应是铝带精轧生产中进行厚度控制的关键,但目前尚不能完全从理论上建立一种普遍适用的数学模型。提出了利用泛化回归RBF神经网络对张力和速度控制系统建模,并根据实际生产数据进行训练和测试,仿真结果表明该模型提高了系统的控制精度,证明了该方法的有效性。
Tensile and velocity effects are the key to thickness control in the production of aluminum strip. However, a universal mathematical model can not be established theoretically at present. The generalized regression RBF neural network is used to model the tension and speed control system. The model is trained and tested according to the actual production data. The simulation results show that the model improves the control precision and proves the effectiveness of the method.