Cutting tool wear classification and detection using multi-sensor signals and Mahalanobis-Taguchi System

Rizal, Muhammad; Ghani, Jaharah A.; Nuawi, Mohd. Zaki; Haron, Che Hassan Che

doi:10.1016/j.wear.2017.02.017

Cited by 76 publications

(33 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The statistical model is also an often experimented approach in TCM systems. Statistical parameters include the root-mean-square, maximum/minimum, average, standard deviation, and kurtosis of time series data [10,11]. Earlier, statistical analysis in fault detection using skewness and kurtosis was vastly prominent.…”

Section: Introductionmentioning

confidence: 99%

Cutting tool wear progression index via signal element variance

Kasim

Nuawi

Ghani

et al. 2019

JMES

Self Cite

View full text Add to dashboard Cite

This paper presents a new statistical-based method of cutting tool wear progression in a milling process called Z-rotation method in association with tool wear progression. The method is a kurtosis-based that calculates the signal element variance from its mean as a measurement index. The measurement index can be implicated to determine the severity of wear. The study was conducted to strengthen the shortage in past studies notably considering signal feature extraction for the disintegration of non-deterministic signals. The Cutting force and vibration signals were measured as a tool of sensing element to study wear on the cutting tool edge at the discrete machining conditions. The monitored flank wear progression by the value of the RZ index, which then outlined in the model data pattern concerning wear and number of samples. Throughout the experimental studies, the index shows a significant degree of nonlinearity that appears in the measured impact. For that reason, the accretion of force components by Z-rotation method has successfully determined the abnormality existed in the signal data for both force and vibration. It corresponds to the number of cutting specifies a strong correlation over wear evolution with the highest correlation coefficient of R2 = 0.8702 and the average value of R2 = 0.8147. The index is more sensitive towards the end of the wear stage compared to the previous methods. Thus, it can be utilised to be the alternative experimental findings for monitoring tool wear progression by using threshold values on certain cutting condition.

show abstract

Section: Introductionmentioning

confidence: 99%

Cutting tool wear progression index via signal element variance

Kasim

Nuawi

Ghani

et al. 2019

JMES

Self Cite

View full text Add to dashboard Cite

show abstract

“…e eigenvalue of the sample covariance matrix becomes equation (8) by processing with the regularization and smoothing techniques. en,…”

Section: Computational Intelligence and Neurosciencementioning

confidence: 99%

“…e MTS is a commonly used multisystem pattern recognition method, which has achieved good results in medical diagnosis [1,2], financial early warning [3], product detection [4,5], fault analysis [6], enterprise management, comprehensive evaluation [7], and so on. e MTS is widely applied to the optimization and classification of large sample data or imbalanced data [6,[8][9][10][11][12]. However, in the field of pattern recognition, a large number of recognition problems belong to high-dimensional small sample size problem, and the research on high-dimensional small sample size problem has gradually become a hot spot.…”

Section: Introductionmentioning

confidence: 99%

Optimized Mahalanobis–Taguchi System for High-Dimensional Small Sample Data Classification

Xiao

Shi

et al. 2020

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

The Mahalanobis–Taguchi system (MTS) is a multivariate data diagnosis and prediction technology, which is widely used to optimize large sample data or unbalanced data, but it is rarely used for high-dimensional small sample data. In this paper, the optimized MTS for the classification of high-dimensional small sample data is discussed from two aspects, namely, the inverse matrix instability of the covariance matrix and the instability of feature selection. Firstly, based on regularization and smoothing techniques, this paper proposes a modified Mahalanobis metric to calculate the Mahalanobis distance, which is aimed at reducing the influence of the inverse matrix instability under small sample conditions. Secondly, the minimum redundancy-maximum relevance (mRMR) algorithm is introduced into the MTS for the instability problem of feature selection. By using the mRMR algorithm and signal-to-noise ratio (SNR), a two-stage feature selection method is proposed: the mRMR algorithm is first used to remove noise and redundant variables; the orthogonal table and SNR are then used to screen the combination of variables that make great contribution to classification. Then, the feasibility and simplicity of the optimized MTS are shown in five datasets from the UCI database. The Mahalanobis distance based on regularization and smoothing techniques (RS-MD) is more robust than the traditional Mahalanobis distance. The two-stage feature selection method improves the effectiveness of feature selection for MTS. Finally, the optimized MTS is applied to email classification of the Spambase dataset. The results show that the optimized MTS outperforms the classical MTS and the other 3 machine learning algorithms.

show abstract

“…La MD es una distancia generalizada que esútil para determinar las similitudes entre conjuntos de datos desconocidos y conocidos. Mide las distancias en espacios multidimensionales; teniendo en cuenta las correlaciones entre cualquier variable o característica que pueda existir [18]. La MD es una medida de distancia que se ha utilizado en aplicaciones como la detección de anomalías, el reconocimiento de patrones y el control de procesos.…”

Section: Distancia Mahalanobis (Md)unclassified

Selección óptima de variables mediante metaheurísticas binarias para la detección de fallas en un proceso industrial multivariado usando aprendizaje de máquina

Acosta¹,

Reséndiz‐Flores²,

García-Calvillo³

2019

RCS

View full text Add to dashboard Cite

Resumen. Este trabajo presenta la comparación entre la optimización por enjambre de partículas y el algoritmo de búsqueda gravitacional en sus versiones binarias en combinación con la distancia Mahalanobis para la selecciónóptima de variables en un proceso de soldadura de discos de soporte para el mecanismo de los asientos en el automóvil. Los resultados muestran que la metodología implementada es capaz de seleccionar un subconjunto de variables capaz de mantener e incluso aumentar la exactitud en la detección de fallas en el proceso, así como brindar a los encargados del proceso las variables que intervienen en la variabilidad del mismo. Lo anterior se valida mediante la implementación de 16 algoritmos y enfoques de aprendizaje automático supervisado y otro no supervisado ampliamente utilizados para tareas de reconocimiento de patrones y detección de fallas.Palabras clave: metaheurísticas binarias, distancia Mahalanobis, aprendizaje automático.Abstract. This work presents the comparison between binary particle swarm optimization and binary gravitational search algorithm in combination with the Mahalanobis distance for the optimal feature selection in a welding process of support discs for the seat mechanism in the car. The results show that the implemented methodology is able to select a subset of variables capable of maintaining and even increasing the accuracy for fault detection in the process. This is validated through 37

show abstract

Cutting tool wear classification and detection using multi-sensor signals and Mahalanobis-Taguchi System

Cited by 76 publications

References 18 publications

Cutting tool wear progression index via signal element variance

Cutting tool wear progression index via signal element variance

Optimized Mahalanobis–Taguchi System for High-Dimensional Small Sample Data Classification

Selección óptima de variables mediante metaheurísticas binarias para la detección de fallas en un proceso industrial multivariado usando aprendizaje de máquina

Contact Info

Product

Resources

About