Using the Improved BP Algorithm of Neural Networks to Predict the Solid Mass Flow of L-Valve in Circulating Fluidized Bed

Wei ZHANG,En-lu WANG,Tian-xing MA,Jian MA
DOI: https://doi.org/10.3321/j.issn:1006-2467.2006.08.013
2006-01-01
Shanghai Jiaotong Daxue Xuebao/Journal of Shanghai Jiaotong University
Abstract:In the basis of improving the traditional BP (back propagation) algorithm of neural networks by using relative error function, regrouping the training sets, adding inertial part, letting parameter change dynamically, and adding random value to the connection matrix, a model of the solid mass flow of L-valve in a circulating fluidized bed with different air charging manner, position, and volume flow was built up by using the improved BP algorithm. The prediction result of the solid mass flow of L-valve through using the improved BP algorithm has a good agreement with the experimental results acquired from the L-valve test facility.
What problem does this paper attempt to address?