基于动态递归神经网络的航煤比重在线估计

摘要

针对常压塔航煤比重在线质量仪表存在滞后大、易出现故障的缺点，提出了运用动态递归神经网络，根据实测的温度、流量、压力等过程参数，在线估计航煤比重，为比重质量闭环控制提供了基础。%In accordance with that on-line specific gravity quality instrument of aviation kerosene in atmospheric distillation tower exists the shortcoming of long time delay and easy trouble, the on line estimation of aviation kerosene specific gravity based on diagonal recurrent neural network is proposed in this paper. The input variables of neural network are temperature, flow, press, which can been measured. The estimation of aviation kerosene specific gravity provides foundation for quality closed control.