Munawar Zaman and Adnan Hassan (2019) . Improved Statistical Feature-based Control Chart Patterns Recognition using ANFIS with Fuzzy Clustering. Neural Computing and Applications, Vol 31 (10). p.5935-5949, Springer. DOI 10.1007/s00521-018-3388-2 (IF: 4.215)
Various types of abnormal control chart patterns can be linked to certain assignable causes in industrial processes. Hence, control chart patterns recognition methods are crucial in identifying process malfunctioning and source of variations. Recently, the hybrid soft computing methods have been implemented to achieve high recognition accuracy. These hybrid methods are complicated, because they require optimizing algorithms. This paper investigates the design of efficient hybrid recognition method for widely investigated eight types of X-bar control chart patterns. The proposed method includes two main parts: the features selection and extraction part and the recognizer design part. In the features selection and extraction part, eight statistical features are proposed as an effective representation of the patterns. In the recognizer design part, an adaptive neuro-fuzzy inference system (ANFIS) along with fuzzy c-mean (FCM) is proposed. Results indicate that the proposed hybrid method (FCM-ANFIS) has a smaller set of features and compact recognizer design without the need of optimizing algorithm. Furthermore, computational results have achieved 99.82% recognition accuracy which is compa- rable to published results in the literature.
Keywords Control chart patterns recognition ? Fuzzy clustering ? ANFIS ? Statistical features
In any quality improvement program, accurate detection of process variations is critical. Normally, such detection is done by using statistical process control (SPC) techniques. For detection of process variations, statistical signals from SPC are used to measure process deviations from stable state. Furthermore, detection of process variations by SPC control chart helps to monitor the process at the required quality level. X-bar control chart developed by Walter A. Shewhart in 1924 remains as one of the most widely used SPC tool to detect process variations in manufacturing .
The behaviors of time-based data plotted on control charts are usually recognized in terms of control chart pat- terns (CCPs). Different types of control chart patterns would be interconnected to some particular assignable causes. Accurately and timely recognition of control chart patterns provides appropriate process monitoring without loss of required quality level. The most widely used control chart patterns are shown in Fig. 1. Brief descriptions of these patterns are given below:
1. Normal pattern (NOR) Normal patterns have points positioned within control limits. Most of the time normal patterns are observed in any system and system is considered as in-statistical control.
2. Trend pattern (TU or TD) A trend pattern is either increasing or decreasing in one direction. The upward movement of points is known as upward trend (TU) and downward movement is called downward trend (TD). Trend patterns could be associated with assign- able causes of tool wear, seasonal effects and operators fatigue, etc.
3. Shift pattern (SU or SD) A shift pattern is the abrupt change in the mean of a process. Upward movement of mean on control chart is called upward shift (SU) and downward movement is called downward shift (SD). Shift patterns are observed when new operator, method or materials are introduced in the systems. Also, if minor failures of machine parts are present shift pattern will be observed.
4. Systematic pattern (SYS) Normal patterns data points behave on control chart haphazard or random. In systematic departure patterns, a high point is always followed by a low point and vice versa. The variation in systematic pattern is high as compare to normal pattern.
5. Cyclic pattern (CYC) Cyclic patterns on control chart can be predicted by a series of high portions or peaks and low portions or trenches. The behavior is observed on control chart like sinusoidal. Cyclic patterns are observed due to systematic seasonal changes like fluctuation in pressure, operator fatigues, and fluctuation in voltages.
6. Stratification pattern (STRA) Stratification behavior on control chart is observed when all data points spread near center line within plus minus one sigma. This unexpected low variation, probably due to computa- tional errors.
Fig. 1 Widely used eight types of control chart patterns
2 ANFIS design and development for CCPs recognizer
3 Results and discussion
4 Conclusions and future work
The extensive practices of online data collection systems have enlarged the need to automate the analysis of process data to minimize process variations with little or no oper- ator involvement. This study presented a new hybrid method for detecting process variations by improving ANFIS performance in two parts; selection and extraction of effective statistical features and development of hybrid recognition method for eight types of widely used patterns. A set of eight statistical features has been selected to achieve better recognition accuracy and to avoid over fit- ting by compact input data set. The new proposed recog- nition method has a combination of fuzzy c-mean (FCM) and ANFIS using eight statistical features. We compared performance of the proposed hybrid FCM–ANFIS method using two types of data sets, randomly generated using standard equations and the published data set. The simu- lation results indicate that the proposed hybrid method achieves comparable classification accuracy for eight types of investigated CCPs without the need of additional optimization algorithm. This study can be extended to further evaluate incomplete and developing patterns.
1. Montgomery DC (2009) Statistical quality control-a modern introduction. Wiley, New York
2. Western Electric Company. (1958) Statistical quality control handbook. The Company
3. Nelson LS (1984) The Shewhart control chart–tests for special causes. J Qual Technol 16(4):237–239
4. Lucy-Bouler TL (1993) Problems in control chart pattern recognition systems. Int J Qual Reliab Manag 10(8):5–13
5. Guh RS (2005) A hybrid learning-based model for on-line detection and analysis of control chart patterns. Comput Ind Eng 49(1):35–62
6. Uguz H (2012) Adaptive neuro-fuzzy inference system for diagnosis of the heart valve diseases using wavelet transform with entropy. Neural Comput Appl 21(7):1617–1628
7. Demircan S, Kahramanli H (2016) Application of fuzzy C-means clustering algorithm to spectral features for emotion classification from speech. Neural Comput Appl. https://doi.org/10.1007/ s00521-016-2712-y
8. Lavanya B, Hannah Inbarani H (2017) A novel hybrid approach based on principal component analysis and tolerance rough similarity for face identification. Neural Comput Appl. https:// doi.org/10.1007/s00521-017-2994-8
9. Kazemi MS, Kazemi K, Yaghoobi MA, Bazargan H (2016) A hybrid method for estimating the process change point using support vector machine and fuzzy statistical clustering. Appl Soft Comput 40:507–516
10. Haykin SS (2001) Neural networks: a comprehensive foundation. Tsinghua University Press, Beijing
11. Vapnik V (2013) The nature of statistical learning theory. Springer, Berlin
12. Ebrahimzadeh A, Ranaee V (2011) High efficient method for control chart patterns recognition. Acta technica Cˇ SAV 56(1):89–101
13. Xanthopoulos P, Razzaghi T (2014) A weighted support vector machine method for control chart pattern recognition. Comput Ind Eng 70:134–149
14. A Viattchenin D, Tati R, Damaratski A (2013) Designing Gaussian membership functions for fuzzy classifier generated by heuristic possibilistic clustering. J Inf Organ Sci 37(2):127–139
15. Zarandi MF, Alaeddini A, Turksen IB (2008) A hybrid fuzzy adaptive sampling–run rules for Shewhart control charts. Inf Sci 178(4):1152–1170
16. Khajehzadeh A, Asady M (2015) Recognition of control chart patterns using adaptive neuro-fuzzy inference system and efficient features. Int J Sci Eng Res 6(9):771–779
17. Khormali A, Addeh J (2016) A novel approach for recognition of control chart patterns: Type-2 fuzzy clustering optimized support vector machine. ISA Trans 63:256–264
18. Das P, Banerjee I (2011) An hybrid detection system of control chart patterns using cascaded SVM and neural network–based detector. Neural Comput Appl 20(2):287–296
19. Pham DT, Wani MA (1997) Feature-based control chart pattern recognition. Int J Prod Res 35(7):1875–1890
- Gauri SK, Chakraborty S (2009) Recognition of control chart patterns using improved selection of features. Comput Ind Eng 56(4):1577–1588
- Hassan A, Baksh MSN, Shaharoun AM, Jamaluddin H (2011) Feature selection for SPC chart pattern recognition using frac- tional factorial experimental design. In: Pham DT, Eldukhri EE, Soroka AJ (eds) Intelligent production machines and system: 2nd I* IPROMS virtual international conference, Elsevier, pp 442–447
- Hassan A, Baksh MSN, Shaharoun AM, Jamaluddin H (2003) Improved SPC chart pattern recognition using statistical features. Int J Prod Res 41(7):1587–1603
- Al-Assaf Y (2004) Recognition of control chart patterns using multi-resolution wavelets analysis and neural networks. Comput Ind Eng 47(1):17–29
- Cheng CS, Huang KK, Chen PW (2015) Recognition of control chart patterns using a neural network-based pattern recognizer with features extracted from correlation analysis. Pattern Anal Appl 18(1):75–86
- Masood I, Hassan A (2010) Issues in development of artificial neural network-based control chart pattern recognition schemes. Eur J Sci Res 39(3):336–355
- Hachicha W, Ghorbel A (2012) A survey of control-chart pattern- recognition literature (1991–2010) based on a new conceptual classification scheme. Comput Ind Eng 63(1):204–222
- Swift JA (1987) Development of a knowledge-based expert system for control-chart pattern recognition and analysis. Okla- homa State Univ, Stillwater.
- De la Torre Gutierrez H, Pham DT (2016) Estimation and gen- eration of training patterns for control chart pattern recognition. Comput Ind Eng 95:72–82
- Alcock (1999) http://archive.ics.uci.edu/ml/databases/synthe ticcontrol/syntheticcontrol.data.html
- Bennasar M, Hicks Y, Setchi R (2015) Feature selection using joint mutual information maximisation. Expert Syst Appl 42(22):8520–8532.
- Haghtalab S, Xanthopoulos P, Madani K (2015) A robust unsu- pervised consensus control chart pattern recognition framework. Expert Syst Appl 42(19):6767–6776
- Bezdek JC, Ehrlich R, Full W (1984) FCM: the fuzzy c-means clustering algorithm. Comput Geosci 10(2–3):191–203
33. Deer PJ, Eklund P (2003) A study of parameter values for a Mahalanobis distance fuzzy classifier. Fuzzy Sets Syst 137(2):191–213
34. Sumathi S, Surekha P, Surekha P (2010) Computational intelli- gence paradigms: theory and applications using MATLAB, vol 1. CRC Press, Boca Raton
35. Jang JS (1993) ANFIS: adaptive-network-based fuzzy inference system. IEEE Trans Syst Man Cybern 23(3):665–685
36. Gacek A, Pedrycz W (eds) (2011) ECG signal processing, clas- sification and interpretation: a comprehensive framework of computational intelligence. Springer, Berlin
37. Joaquim PMDS, Marques S (2007) Applied statistics using SPSS, statistica, Matlab and R. Springer, Berlin, pp 205–211
38. Carbonneau R, Laframboise K, Vahidov R (2008) Application of machine learning techniques for supply chain demand forecast- ing. Eur J Oper Res 184(3):1140–1154
39. Stukowski A (2009) Visualization and analysis of atomistic simulation data with OVITO—the Open Visualization Tool. Modell Simul Mater Sci Eng 18(1):015012
40. Mylonopoulos NA, Doukidis GI, Giaglis GM (1995). Assessing the expected benefits of electronic data interchange through simulation modelling techniques. In: The proceedings of the 3rd European conference on information systems, Athens, Greece, pp 931–943
41. Abdulmalek FA, Rajgopal J (2007) Analyzing the benefits of lean manufacturing and value stream mapping via simulation: a pro- cess sector case study. Int J Prod Econ 107(1):223–236
42. Assaleh K, Al-Assaf Y (2005) Features extraction and analysis for classifying causable patterns in control charts. Comput Ind Eng 49(1):168–181