平整机轧制力的预报对轧制过程的优化控制有着重要意义。针对平整机轧制力预测精度不高的问题,提出采用Re LU(Rectified Linear Units)激活函数的神经网络模型来预报平整机的轧制力。在对数据进行主成分分析后,得到影响轧制力的主要因素,并将其作为神经网络的输入层,将平整机轧制力作为输出层,通过使用Python语言编程进行实验,对神经网络模型隐层的相关参数及算法进行单一变量筛选,建立了保证轧制力预报精度最高的神经网络模型。实验结果表明,通过调整隐层层数、神经元数、传播算法、正则化方法,该模型能够将预测误差控制在10%以内,且该实验方法能够对不同输入参数下的平整机轧制力进行精确预报。 The forecast of the rolling force of the flattening machine is of great significance for the optimization and control of the rolling process. Aiming at the problem that the prediction accuracy of the rolling force of the flattening machine is not high, a neural network model of Re LU (Rectified Linear Units) activation function is proposed to predict the rolling force of the flattening machine. After the principal component analysis of the data, the main factors influencing the rolling force are obtained and used as the input layer of the neural network, and the rolling force of the flattening machine is taken as the output layer. Experiments are conducted by using the Python language programming, and the neural network Model hidden layer parameters and algorithms for a single variable screening, established to ensure the accuracy of rolling force prediction neural network model. The experimental results show that the model can control the prediction error within 10% by adjusting the number of hidden layers, the number of neurons, the propagation algorithm and the regularization method. The experimental method can simulate the rolling of flattening machine under different input parameters Force for accurate forecast.