Abstract: To improve the measurement accuracy of coal powder mass flow measurement in industrial production, based on the double-elbow method of measuring mass flow, a radial basis function network soft measurement model was established using the powerful nonlinear mapping ability of artificial neural network and reflecting difficult determined influence factors of solid mass flow to network connection weights. The experiment was conducted on the coal powder pneumatic conveying experiment platform to get experimental data. The model was trained and simulated taking experimental data as samples. The results show that the measure error of the model is within 3% which provides a simple and effective method for solid phase mass flow measurement of coal powders.

Keywords: coal powders; mass flow; double-elbow method; radial basis function network

中图分类号：TB937        文献标志码：A

文章编号：1008-5548（2014）03-0067-03

DOI:10.13732/j.issn.1008-5548.2014.03.016

收稿日期：2013-11-28， 修回日期：2013-12-19，在线出版时间：2014-06-24。

基金项目：国家自然科学基金项目，编号：61271402；河北省自然科学基金项目，编号：F2010001970。

第一作者简介：赵延军(1972—)，男，博士，副教授，研究方向为气-固两相流参数测量及 CEMS 系统。 E-mail：zhyj_ts@sohu.com。

通信作者简介：程守光(1986—)，男，硕士研究生，研究方向为气-固两相流参数测量。 E-mail：chengshouguang@126.com。