A prediction method for intervals of trace ions concentration in zinc sulfate solution based on UV-vis spectroscopy
Abstract
The acquisition of intervals of trace ions concentration based on spectrophotometry is an essential stage in hydrometallurgical purification process. The spectral signals of ions have the characteristics of skewed sample distribution and unequal misjudgment costs. Thus, a novel objective optimization method is proposed for interval prediction in this study. First, wavelet transform is utilized to obtain derivative denoising spectrum for zinc sulfate solution spectral signal, and a feature extraction method is used to reduce multicollinearity and identify key factors. Second, the optimization objective of support vector machine (SVM) model is defined to solve skewed sample distribution and unequal misjudgment costs. At last, the optimal parameters of the model are solved by stochastic search algorithm. A spectra dataset containing 64 samples of high concentration Zn2+, trace Co2+ and Cu2+ mixture solution is prepared for the proposed method. Experimental results have shown the feasibility and effectiveness of the proposed method.
- Publication:
-
Optik
- Pub Date:
- October 2019
- DOI:
- 10.1016/j.ijleo.2019.163065
- Bibcode:
- 2019Optik.19463065Z
- Keywords:
-
- UV-vis spectrophotometry;
- Concentration interval;
- Trace impurities;
- Objective optimization;
- Support vector machine