【摘要】 采用弯管法测量稀相气固两相流中固相质量流量时,固相流量与其影响因素(压差、流量系数、气固混合密度等)之间存在着复杂的非线性关系,给粉体的精确测量带来困难。利用人工神经网络优良的非线性映射能力,建立了一个基于BP网络固相质量流量软测量模型,并以实验数据为样本对网络进行训练,实现对固相质量流量的在线估计,与实验结果吻合较好,为稀相气力输送中固相质量流量在线测量提供了一种简单、可靠的新方法。更多还原

【Abstract】 It is very difficult for the elbow method to accurately measure the mass flow rate of pneumatically conveyed solids due to the complex nonlinear relation between solid mass flow rate and its effects such as differential pressure, flow coefficient and bulk density,etc.Based on better nonlinear approximation capability of artificial neural network,a soft sensor model is introduced to realize the above nonlinear relation and to provide a solution to on-line measurement of pneumatically conveyed solids.Experimental results obtained on a pilot gas-solid conveyor showed a good agreement with those of soft sensor,which proves the validity and reliability of the soft sensor.更多还原