论文部分内容阅读
为准确预测双循环流化床生物质气化的颗粒循环流率Gs,设计并搭建了双循环流化床冷态试验台,研究了提升管流化风速、二次风量、二次风送风方式、二次风口高度及二次风口数量对颗粒循环流率的影响,并建立了基于Levenberg-Marquardt优化算法的BP神经网络预测模型,通过对比找出了最优模型,对颗粒循环流率进行了预测.结果表明:Gs随着提升管流化风速和二次风量的增大而增加,二次风量超过一定值后,增加的趋势变缓;二次风径向引入比切向引入时的Gs大;Gs对二次风口高度的变化十分敏感;应用该BP神经网络模型得出的Gs预测值与试验值的平均偏差为0.07 kg/(m2.s),平均相对误差仅为0.49%,模型准确地预测了提升管送风特性对颗粒循环流率的影响.
In order to accurately predict the circulating flow rate Gs of double-circulating fluidized bed biomass gasification, a double-circulating fluidized bed cold-state test bench was designed and constructed to study the effect of riser fluidization velocity, secondary air flow, secondary air supply Mode, the height of the secondary tuyere and the number of secondary tuyeres on the circulating flow rate of the particles, and the BP neural network prediction model based on the Levenberg-Marquardt optimization algorithm was established. By comparing and finding the optimal model, the particle circulation flow rate The results show that Gs increases with the increase of the fluidized velocity and secondary air volume of the riser, and the tendency of increase increases with the secondary air volume exceeding a certain value. Gs is large; Gs is very sensitive to the change of secondary tuyere height. The average deviation of Gs forecast value and test value obtained by this BP neural network model is 0.07 kg / (m2.s), the average relative error is only 0.49% The model accurately predicts the effect of riser air distribution on the particle flow rate.