摘要
在水平T形分支管道中,用压缩空气作为输送气体,对不同粒径的砂石进行气力输送试验。分别通过试验和改进型BP神经网络预测两种方法对表观气速和分支管路流量控制阀开度变化时,固相在分支管路中的分配特性进行了研究。结果表明,随着表观气速减小和两分支管路流量控制阀开度差值变大,固相流量在两分支管中的分配产生较大差异。试验值和改进型BP网络预测值的对比结果表明,二者相互吻合较好,说明采用改进型的BP网络来模拟固相在分支管路中的分配特性适应性较好。
Pneumatic conveying experiments of different particle diameter sands were carried out using compress air at horizontal tee branch pipe. The paper investigated the flow distribution of solid phase for branch pipe through experiments and improved BP neural network when both conveying velocity and the flow valve opening of branch pipe were changed. When the conveying velocity is decreased and the flow valve opening of branch pipe is visible difference, it is found that the flow distribution ratio of solid phase in the branch pipe has some obvious changes. With the comparison between the neural network simulation results and the experimental data, it is shown that the simulation values agree well with the experimental data. This illuminates that the improved BP network has a better adjustability to the simulation and prediction of the flow distribution characteristics of solid phase for tee branch pipe.
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2006年第20期2110-2112,共3页
China Mechanical Engineering
基金
上海市科学技术委员会资助项目(03JC14055)
上海市教育委员会资助项目(05EZ15)
关键词
气力输送
分支管
流量分配
神经网络
pneumatic conveying
branch pipe
flow distribution
neural network