Affiliation:
1. East China University of Science and Technology
Abstract
A 17-27-5 type BP neural network model was built, whose sampled data was got by hydrocyclone separation experiments; another 6-30-5 type BP neural network was also built, whose sampled data came from the simulation results of the LZVV of a hydrocyclone with CFD code FLUENT. The two neural network models also have good predictive validity aimed at hydrocyclone separation performance. It demonstrates LZVV structural parameters can embody hydrocyclone separation performance and reduce input parameter numbers of neural network model. It also indicates that the predictive model of hydrocyclone separation performance can be built by neural network.
Publisher
Trans Tech Publications, Ltd.
Reference7 articles.
1. Y. H. Zhang, Y. Liu, P. Qian and H. L. Wang, Experimental Investigation of a Minihydrocyclone, Chemical Engineering & Technology, Vol. 32, No. 8 (2009), pp.1274-1279.
2. Z. S. Bai, H. L. Wang and S. T. Tu, Numerical and Experimental Study on the Removal of Catalyst Particles from Oil Slurry by Hydrocyclone, Petroleum Science and Technology, Vol. 28, No. 5 (2010), pp.525-533.
3. L. Y. Chu, W. M. Chen and X. Z. Lee, Enhancement of hydrocyclone performance by controlling the inside turbulence structure, Chemical Engineering Science, Vol. 57, No. 1 (2002), pp.207-212.
4. C. L. Karr, D. A. Stanley and B. McWhorter, Optimization of hydrocyclone operation using a gene-fuzzy algorithm, Computer Methods in Applied Mechanics and Engineering, Vol. 186, No. 2-4 (2000), pp.517-530.
5. T. Q. Wang, S. J. Wang and H. X. Zhang, Dynamic Extreme Learning Machine: A Learning Algorithm for Neural Network with Elastic Output Structure (2009), pp.271-275.
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献