Soft-sensing Model of SAG Mill Based on PSPCAW-SVM

By analyzing the soft-sensing modeling process of SAG based on support vector machine, the problem mainly involves the selection of auxiliary variables and data processing in soft-sensing modeling. In this paper, the correlation coefficient method is used to analyze the correlation between the dominant variable and the auxiliary variable, and the auxiliary variable with high correlation is selected. The phase space of the signal related to the load of SAG is reconstructed, and the selection dimension of the auxiliary variables is preliminarily determined, and the principal component analysis technology is used to fuse some auxiliary variables to prevent information redundancy. Then it combines the three techniques to construct an improved PSPCA algorithm for screening auxiliary variables, and determines the auxiliary variables of the soft-sensing model, and wavelet denoising filtering is performed on the data sample of the determined auxiliary variables to remove the noise in the original data. Finally, PSPCAW-SVM soft-sensing model is established to lay the foundation for further research.