论文部分内容阅读
在锥形布风板双循环流化床冷态装置上,研究了提升管风速、气化室风速、物料质量和颗粒粒径对提升管颗粒循环流率的影响,并与水平布风板的结果进行了对比.利用3种改进的BP神经网络算法建立模型来预测循环流率.结果表明:提升管颗粒循环流率随着提升管风速和气化室风速的增大而增大,当风速达到一定值后,增大趋势逐渐平缓;循环流率随着物料质量的增大基本呈线性增大,随着颗粒粒径的增大而明显减小;锥形布风板比水平布风板更具优势,同样条件下可以增大循环流率;BFGS拟牛顿算法的预测效果最佳,其颗粒循环流率预测值与实验值的最大相对误差为7.703 5%,平均相对误差为3.594 3%.
The effect of the velocity of the riser, the velocity of the gasification chamber, the mass of the gasification chamber and the particle size on the circulation flow rate of the riser was studied on a double-circulating fluidized bed cold-state apparatus with tapered baffles. The results were compared.The three models of BP neural network algorithm were used to predict the circulating flow rate.The results show that the circulating flow rate of the riser increases with the increase of the riser tube velocity and the gas velocity in the gasification chamber, After a certain value, the trend of increasing gradually flatten; the circulating flow rate increases linearly with the increase of material quality, decreases obviously with the increase of particle size; Which has the advantage of increasing the circulating flow rate under the same conditions. The predicted result of BFGS quasi-Newton algorithm is the best. The maximum relative error between predicted and experimental values of particle circulation flow rate is 7.703 5% and the average relative error is 3.594 3%.