论文部分内容阅读
利用现场的数据 ,采用 BP神经元网络预报热连轧层流水冷区集管组内的基本热流密度 ,将预报的结果用于上、下集管组的热流密度的数学模型计算 ,进而优化层冷集管组的水冷温降计算数学模型的精度。将结果与采用多元回归方法所得到的结果作比较 ,表明采用 BP神经元网络计算基本热流密度的精度要高于多元回归方法的计算精度 ,卷取温度的计算值与实测值的标准差比解析回归方法减少了近 2 0 % ,说明该方法具有良好的在线应用前景。
Based on the field data, the BP neural network is used to predict the basic heat flux density in the header zone of the hot-strip continuous flow cooling zone. The prediction results are used to calculate the heat flux density in the upper and lower headers. Accuracy of mathematic model for cold water temperature drop calculation in cold header. Comparing the result with the result obtained by multivariate regression method, it shows that the accuracy of using BP neural network to calculate the basic heat flux density is higher than that of multivariate regression method. The standard deviation ratio between calculated and measured values of coiling temperature The regression method is reduced by nearly 20%, indicating that this method has a good online application prospect.