Modeling Leaching Percentage of Zinc-oxide-ore Based on Radial Basis Function Networks and Principal Component Regression
There is complex nonlinear relationship between the leaching percentage of zinc-oxide-ore and its factors, so it is difficult to model by its mechanism. In this paper, a RBFN-PCR approach which radial basis function networks combined with principal component regression was proposed to model for the leaching percentage of zinc-oxide-ore. Cross-validating indicated that the performances of match and predicting of the model were satisfied. The average match relative error and average predicting relative error was 2.55% and 5.07%, respectively.
zinc-oxide-ore leaching percentage radial basis function network principal component regression
ZHENG Qifu WU Guoquan
Department of Chemical Engineering, West Branch of Zhejiang University of Technology, P.R.China, 324 Department of Mathematics and Physics, West Branch of Zhejiang University of Technology, P.R.China,
国际会议
威海
英文
106-109
2010-07-24(万方平台首次上网日期,不代表论文的发表时间)