作　　者： ; ; ; ; ; ;

机构地区： 燕山大学

出　　处： 《中国机械工程》 2005年第11期990-992,共3页

摘　　要： 热轧带钢卷取温度是影响成品带钢性能指标的重要工艺参数之一,其层流控制系统具有高度的非线性。采用附加动量BP算法,建立了基于神经网络前馈与数学模型反馈的联合层流控制系统,仿真结果表明,采用神经网络预测的卷取温度与实测温度相近,结果可信,为层流数学模型参数的在线辨识打下了坚实的基础。 Hot strip coiling temperature is one of the important parameters of performance index in hot rolled strip, and its control systems of highly nonlinearity. The coiling temperature of hot rolled strip is exactly predicted based on neural network (NN) and an improved BP algorithm. A new coiling temperature system based on NN combined with mathematical model was presented, where the feed-forward control was based on the NN and the feed-backward control was based on mathematic model. Finally, the prediction by the NN shows that the control performance is satisfactory, and it can make mathematical model of coiling temperature identified.

关 键 词： 热轧带钢 层流冷却 卷取温度 神经网络 算法

领　　域： [金属学及工艺]