基于改进BP算法对双循环流化床颗粒流率预测研究

Prediction Research on Particle Rate in a Dual Circulating Fluidized Bed Based on Improved BP Algorithm

锥形布风板导致径向物料厚度不同,进而造成不均匀布风,增强颗粒扰动。合理地控制颗粒循环流率是锥形布风板双流化床生物质气化装置稳定运行的关键。在自行搭建的冷态装置上,针对气化室风速、提升管风速、物料重量、物料粒径以及锥形布风板角度对颗粒流率的影响进行了实验研究。结果表明:颗粒循环流率随着气化室风速、提升管风速、物料重量的增加而增加,随着物料粒径的增大而明显减小。利用matlab建立了LM-BP神经网络模型来预测循环流率,通过对比找出了最优模型,最大相对误差为7.65%,平均相对误差为2.5317%。

The conical distributor can bring about different radial material thickness, cause uneven air distribution and enhance particle disturbance. The key to keep dual fluidized bed biomass gasifier with a cone air distributor running steadily is to control the particle circulation rate reasonably. On a cold device, how wind speed of the gasification chamber, riser＇s wind speed, weight of the material, particle size and distributor＇s angle affect particle flow rate were experimental studied. The study show that： the particle circulation flow rate increases with the increasing of wind speed in the gasification chamber or riser and material weight, significantly reduces with increasing particle size. Using matlab to build LM-BP neural networks to predict the circulation flow rate. The best model is found by comparison： the maximum relative error is 7.65% and the mean relative error is 2.5317%.