The computational speed in the feature selection of Mahalanobis-Taguchi system (MTS) using standardbinary particle swarm optimization (BPSO) is slow and it is easy to fall into the locally optimal solution. Thispaper proposes an MTS variable optimization method based on chaos quantum-behavior particle swarm. Inorder to avoid the influence of complex collinearity on the distance measurement results, the Gram-Schmidtorthogonalization method is first used to calculate the Mahalanobis distance (MD) value. Then, the optimalthreshold point of the system classification is determined through the receiver operating characteristic (ROC)curve; the misclassification rate and the selected variables are defined; the multi-objective mixed programmingmodel is built. The chaos quantum-behavior particle swarm optimization (CQPSO) algorithm is proposed to solvethe optimization combination, and the algorithm performs binary coding on the particle based on probability.Using the optimized combination of variables, a new Mahalanobis-Taguchi metric based prediction system isestablished to complete the task of precise discrimination. Finally, a fault diagnosis for the steel plate is taken asan example. The experimental results show that the proposed method can effectively enhance the iterative speedand optimization precision of the particles, and the prediction accuracy of the optimized MTS is significantlyimproved.
LIU Jiufu ∗ (刘久富), ZHENG Rui (郑锐), ZHOU Zaihong ∗ (周再红), ZHANG Xinzhe (张信哲),YANG Zhong (杨忠), WANG Zhisheng (王志胜)
. Feature Selection Optimization for Mahalanobis-Taguchi SystemUsing Chaos Quantum-Behavior Particle Swarm[J]. Journal of Shanghai Jiaotong University(Science), 2021
, 26(6)
: 840
-846
.
DOI: 10.1007/s12204-020-2236-6
[1] YANG T H, CHENG Y T. The use of Mahalanobis-Taguchi system to improve flip-chip bumping heightinspection efficiency [J]. Microelectronics Reliability,2010, 50(3): 407-414.
[2] HUANG C L, HSU T S, LIU C M. The Mahalanobis-Taguchi system – neural network algorithm for dataminingin dynamic environments [J]. Expert Systemswith Applications, 2009, 36(3): 5475-5480.
[3] SHAKYA P, KULKARNI M S, DARPE A K. Bearingdiagnosis based on Mahalanobis-Taguchi-Gram-Schmidt method [J]. Journal of Sound and Vibration,2015, 337: 342-362.
[4] RIZAL M, GHANI J A, NUAWI M Z, et al. Cuttingtool wear classification and detection using multisensorsignals and Mahalanobis-Taguchi system [J].Wear, 2017, 376(15): 1759-1765.
[5] YAZID A M, RIJAL J K, AWALUDDIN M S, etal. Pattern recognition on remanufacturing automotivecomponent as support decision making usingMahalanobis-Taguchi system [J]. Procedia CIRP, 2015,26: 258-263.
[6] ZENG J H, ZENG F Z. The measurement scale ofMahalanobis-Taguchi system optimization based onfuzzy robustness discriminant criterion [J]. IndustrialEngineering and Management, 2008, 13(3): 52-55 (inChinese).
[7] IQUEBAL A S, PAL A, CEGLAREK D, et al. Enhancementof Mahalanobis-Taguchi system via roughsets based feature selection [J]. Expert Systems withApplications, 2014, 41(17): 8003-8015.
[8] NIU J L. Methods of classification and sort evaluationusing Mahalanobis-Taguchi system based onomni-optimizer algorithm and applications [D]. Nanjing:Nanjing University of Science & Technology,2012.
[9] RES′ ENDIZ E, MONCAYO-MART′INEZ L A, SOL′ISG. Binary ant colony optimization applied to variablescreening in the Mahalanobis-Taguchi system [J]. ExpertSystems with Applications, 2013, 40(2): 634-637.
[10] PAL A, MAITI J. Development of a hybrid methodologyfor dimensionality reduction in Mahalanobis-Taguchi system usingMahalanobis distance and binaryparticle swarm optimization [J]. Expert Systems withApplications, 2010, 37(2): 1286-1293.
[11] JIN X H, CHOWTWS. Anomaly detection of coolingfan and fault classification of induction motor usingMahalanobis-Taguchi system [J]. Expert Systems withApplications, 2013, 40(15): 5787-5795.
[12] LIPARAS D, LASKARIS N, ANGELIS L. Incorporatingresting state dynamics in the analysis of encephalographicresponses by means of the Mahalanobis-Taguchi strategy [J]. Expert Systems with Applications,2013, 40(7): 2621-2630.
[13] MAHALAKSHMI P, GANESAN K. MahalanobisTaguchi system based criteria selection for shrimpaquaculture development [J]. Computers and Electronicsin Agriculture, 2009, 65(2): 192-197.
[14] XI M L, SUN J, WU Y. Quantum-behaved particleswarm optimization with binary encoding [J]. Controland Decision, 2010, 25(1): 99-104.
