Data driven particle size estimation of hematite grinding process using stochastic configuration network with robust technique

Abstract

As a production quality index of hematite grinding process, particle size (PS) is hard to be measured in real time. To achieve the PS estimation, this paper proposes a novel data driven model of PS using stochastic configuration network (SCN) with robust technique, namely, robust SCN (RSCN). Firstly, this paper proves the universal approximation property of RSCN with weighted least squares technique. Secondly, three robust algorithms are presented by employing M-estimation with Huber loss function, M-estimation with interquartile range (IQR) and nonparametric kernel density estimation (NKDE) function respectively to set the penalty weight. Comparison experiments are first carried out based on the UCI standard data sets to verify the effectiveness of these methods, and then the data-driven PS model based on the robust algorithms are established and verified. Experimental results show that the RSCN has an excellent performance for the PS estimation.

Keywords

hematite grinding process / particle size / stochastic configuration network / robust technique / M-estimation / nonparametric kernel density estimation

References

[1] ChenX-s, LiQ, FeiS-min. Supervisory expert control for ball mill grinding circuits [J]. Expert Systems with Applications, 2008, 34(3): 1877-1885

[2] ZhouP, DaiW, ChaiT-you. Multivariable disturbance observer based advanced feedback control design and its application to a grinding circuit [J]. IEEE Transactions on Control Systems Technology, 2014, 22(4): 1474-1485

[3] ShenL, HeJ-j, YuS-y, GuiW-hua. Temperature control for thermal treatment of aluminum alloy in a large-scale vertical quench furnace [J]. Journal of Central South University, 2016, 23(7): 1719-1728

[4] QiaoJ-h, ChaiT-you. Soft measurement model and its application in raw meal calcination process [J]. Journal of Process Control, 2012, 22(1): 344-351

[5] YuanZ-t, LiL-x, HanY-x, LiuL, LiuTing. Fragmentation mechanism of low-grade hematite ore in a high pressure grinding roll [J]. Journal of Central South University, 2016, 23(11): 2838-2844

[6] WangX-l, GuiW-h, YangC-h, WangY-lin. Wet grindability of an industrial ore and its breakage parameters estimation using population balances [J]. International Journal of Mineral Processing, 2011, 98(12): 113-117

[7] SunZ, WangH-g, ZhangZ-pu. Soft sensing of overflow particle size distributions in hydrocyclones using a combined method [J]. Tsinghua Science & Technology, 2008, 13(1): 47-53

[8] WangX-h, GuiW-h, WangY-l, YangC-hua. Prediction model of grinding particle size based on hybrid kernel function support vector machine [J]. Computer Engineering and Applications, 2010, 46(2): 207-209

[9] AldrichC, MaraisC, SheanB J, CilliersJJ. Online monitoring and control of froth flotation systems with machine vision: A review [J]. International Journal of Mineral Processing, 2010, 96(1): 1-13

[10] VillarR G D, ThibaultJ, VillarR D. Development of a softsensor for particle size monitoring [J]. Minerals Engineering, 1996, 9(1): 55-72

[11] WangD, LiuJ, SrinivasanR. Data-driven soft-sensor approach for quality prediction in a refining process [J]. IEEE Transactions on Industrial Informatics, 2010, 6(1): 11-17

[12] SbarbaroD, AscencioP, EspinozaP, MujicaF, CortesG. Adaptive soft-sensors for on-line particle size estimation in wet grinding circuits [J]. Control Engineering Practice, 2008, 16(2): 171-178

[13] MarzbanradJ, MashadiB, AfkarA, PahlavaniM. Dynamic rupture and crushing of an extruded tube using artificial neural network (ANN) approximation method [J]. Journal of Central South University, 2016, 23(4): 869-879

[14] KoY D, ShangH. Time delay neural network modeling for particle size in SAG mills [J]. Powder technology, 2011, 205(1): 250-262

[15] LiuJ-p, NiuD-x, ZhangH-y, WangG-qing. Forecasting of wind velocity: an improved svm algorithm combined with simulated annealing [J]. Journal of Central South University, 2013, 20(2): 451-456

[16] PaoY H, TakefujiY. Functional-link net computing: theory, system architecture, and functionalities [J]. Computer, 1992, 25(5): 76-79

[17] IgelnikB, PaoY H. Stochastic choice of basis functions in adaptive function approximation and the functional-link net [J]. IEEE Transactions on Neural Networks, 1995, 6(6): 1320-1329

[18] ScardapaneS, WangD-hui. Randomness in neural networks: An overview [J]. Wiley Interdisciplinary Reviews: Data Mining & Knowledge Discovery, 2017, 7(2): 1-18

[19] ReedR. Pruning algorithms—A survey [J]. IEEE transactions on Neural Networks, 1993, 4(5): 740-747

[20] KwokT Y, YeungD Y. Objective functions for training new hidden units in constructive neural networks [J]. IEEE Transactions on Neural Networks, 1997, 8(5): 1131-1148

[21] KwokT Y, YeungD Y. Constructive algorithms for structure learning in feedforward neural networks for regression problems [J]. IEEE Transactions on Neural Networks, 1996, 8(3): 630-645

[22] WeigendA S, RumelhartD E, HubermanB ALippmannR, MoodyJ, TouretzkyD S. Generalization by weight-elimination with application to forecasting [C]. Advances in Neural Information Processing Systems, 1991, Morgan Kaufmann, San Mateo, CA: 875882

[23] LiM, WangD-hui. Insights into randomized algorithms for neural networks: Practical issues and common pitfalls [J]. Information Sciences, 2017, 382–383: 170-178

[24] WangD-h, LiMing. Stochastic configuration networks: Fundamentals and algorithms [J]. IEEE Transactions on Cybernetics, 2017, 47(10): 3466-3479

[25] WangD-h, LiMing. Robust stochastic configuration networks with kernel density estimation for uncertain data regression [J]. Information Sciences, 2017, 412–413: 210-222

[26] PlittL R. A mathematical model of the gravity classifier [C]. Proceedings of 17th International Mineral Processing Congress. Munich, 1991123135

[27] BondF C. Crushing and grinding calculations, Part I [J]. British Chemical Engineering, 1961, 6(6): 378-385

[28] DaiW, ChaiT-y, YangS X. Data-driven optimization control for safety operation of hematite grinding process [J]. IEEE Transactions on Industrial Electronics, 2015, 62(5): 2930-2941

[29] RousseeuwP J, CrouxC. The bias of k-step M-estimators [J]. Statistics & Probability Letters, 1994, 20(5): 411-420

[30] FanJ, YanA-l, XiuN-hua. Asymptotic properties for M-estimators in linear models with dependent random errors [J]. Journal of Statistical Planning & Inference, 2014, 148(148): 49-66

[31] ScottD WMultivariate density estimation: Theory, practice, and visualization [M], 2015, John Wiley and Sons, Houston, Texa

[32] BLAKE C L, MERZ C J. UCI repository of machine learning databases, [DB/OL]. [1998]. http://www.ics.uci.edu/~mlearn/MLRepository.html.

[33] DaiW, ZhouP, ZhaoD-y, LuS-w, ChaiT-you. Hardware-in-the-loop simulation platform for supervisory control of mineral grinding process [J]. Powder Technology, 2016, 288: 422-434