Understanding ART-based neural algorithms as statistical tools for manufacturing process quality control

Pacella, Massimo; Semeraro, Quirico

doi:10.1016/j.engappai.2005.02.001

Cited by 13 publications

(11 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Among these, Al‐Ghanim presented an adaptive resonance theory (ART) neural network to distinguish natural from unnatural variations in the outcomes of a manufacturing process. Pacella and Semeraro extended the study of Al‐Ghanim by proposing a fuzzy ART neural network to address arbitrary sequences of input patterns, whereas the ANN model proposed by Al‐Ghanim was limited to binary inputs. Pacella and Semeraro derived a variant of the previously proposed fuzzy‐ART‐based scheme to monitor the stability over time of profile data.…”

Section: A Rationale For the Use Of One‐class‐classification Methodsmentioning

confidence: 99%

“…Their performances are evaluated using Monte Carlo simulations in the presence of multimode data and a real dataset acquired in roll grinding operations. The improved K ‐chart design proposed by Ning and Tsung and the fuzzy‐ART‐based scheme proposed by Pacella and Semeraro are reviewed and compared. The previous work of Pacella and Semeraro focused on process monitoring of univariate time series or streams of profile data.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Comparison Study of Distribution‐Free Multivariate SPC Methods for Multimode Data

Grasso

Colosimo

Semeraro

et al. 2014

Quality & Reliability Eng

Self Cite

View full text Add to dashboard Cite

The data-rich environments of industrial applications lead to large amounts of correlated quality characteristics that are monitored using Multivariate Statistical Process Control (MSPC) tools. These variables usually represent heterogeneous quantities that originate from one or multiple sensors and are acquired with different sampling parameters. In this framework, any assumptions relative to the underlying statistical distribution may not be appropriate, and conventional MSPC methods may deliver unacceptable performances. In addition, in many practical applications, the process switches from one operating mode to a different one, leading to a stream of multimode data.Various nonparametric approaches have been proposed for the design of multivariate control charts, but the monitoring of multimode processes remains a challenge for most of them. In this study, we investigate the use of distribution-free MSPC methods based on statistical learning tools. In this work, we compared the kernel distance-based control chart (K-chart) based on a one-class-classification variant of support vector machines and a fuzzy neural network method based on the adaptive resonance theory. The performances of the two methods were evaluated using both Monte Carlo simulations and real industrial data. The simulated scenarios include different types of out-of-control conditions to highlight the advantages and disadvantages of the two methods. Real data acquired during a roll grinding process provide a framework for the assessment of the practical applicability of these methods in multimode industrial applications.

show abstract

Section: A Rationale For the Use Of One‐class‐classification Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A Comparison Study of Distribution‐Free Multivariate SPC Methods for Multimode Data

Grasso

Colosimo

Semeraro

et al. 2014

Quality & Reliability Eng

Self Cite

View full text Add to dashboard Cite

show abstract

“…This drawback can be mainly ascribed to the binary coding of the ART algorithm as it is a less flexible way of using process data than a method based on graded continuous number encoding. In fact, subsequent researches extended this methodology and presented outperforming ART neural networks for unnatural behavior detection (Pacella & Semeraro, 2005;Pacella, Semeraro, & Anglani, 2004a, 2004b. In particular, a simplified Fuzzy ART algorithm, which incorporates computations from fuzzy set theory by which can categorize analog patterns, was presented.…”

Section: Art Neural Network For Process Monitoringmentioning

confidence: 98%

“…In Pacella et al (2004b) it was demonstrated that the training set can even be limited to a single vector whose components are equal to the process nominal value. Fuzzy ART responses to inputs can be easily explained (Pacella & Semeraro, 2005), in contrast to other neural networks, where typically it is more difficult to realize why an input produces a specific output.…”

Section: Art Neural Network For Process Monitoringmentioning

confidence: 99%

Monitoring roundness profiles based on an unsupervised neural network algorithm

Pacella

Semeraro

2011

Computers & Industrial Engineering

Self Cite

View full text Add to dashboard Cite

“…The use of a quality control system can lead to the elimination of assignable causes pointed to by unnatural behaviour 2 . FMEA, providing a framework for cause and effect analysis of potential product or process failures 3 , is a widely used engineering technique for designing, identifying and eliminating known and/or potential failures, problems, errors and so on from system, design, process, and/or service before they reach the customer 4 .…”

Section: Introductionmentioning

confidence: 99%

A Fuzzy Hybrid Approach for Fuzzy Process FMEA: An Application to a Spindle Manufacturing Process

Ekmekçioğlu¹,

Kutlu²

2012

IJCIS

View full text Add to dashboard Cite

Process Failure Modes and Effects Analysis (PFMEA) concept, has been developed based on the success of Failure Modes and Effects Analysis (FMEA) to include a broader analysis team for the realization of a comprehensive analysis in a short time. The most common use of the PFMEA involves manufacturing processes as they are required to be closely examined against any unnatural deviation in the state of the process for producing products with consistent quality. In a typical FMEA, for each failure modes, three risk factors; severity (S), occurrence (O), and detectability (D) are evaluated and their multiplication derives the risk priority number (RPN). However there are many shortcomings of this classical crisp RPN calculation. This study introduces a fuzzy hybrid approach that allows experts to use linguistic variables for determining S, O, and D for PFMEA by applying fuzzy 'technique for order preference by similarity to ideal solution' (TOPSIS) and fuzzy 'analytical hierarchy process' (AHP). An application to a spindle manufacturing process expresses the relevance of the fuzzy hybrid model in PFMEA.

show abstract

Understanding ART-based neural algorithms as statistical tools for manufacturing process quality control

Cited by 13 publications

References 33 publications

A Comparison Study of Distribution‐Free Multivariate SPC Methods for Multimode Data

A Comparison Study of Distribution‐Free Multivariate SPC Methods for Multimode Data

Monitoring roundness profiles based on an unsupervised neural network algorithm

A Fuzzy Hybrid Approach for Fuzzy Process FMEA: An Application to a Spindle Manufacturing Process

Contact Info

Product

Resources

About