• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.15 no.1
• /
• pp.95-102
• /
• 2016

Wrinkle, spring-back, and fracture are major defects frequently found in the sheet metal forming process, and the reduction of such defects is difficult as they are affected by uncontrollable factors, such as variations in properties of the incoming material and process parameters. Without any countermeasures against these issues, attempts to reduce defects through optimal design methods often lead to failure. In this research, a new multi-attribute robust design methodology, based on the Mahalanobis Taguchi System (MTS), is presented for reducing the possibilities of wrinkle, spring-back, and fracture. MTS performs experimentation, based on the orthogonal array under various noise conditions, uses the SN ratio of the Mahalanobis distance as a performance metric. The proposed method is illustrated through a robust design of the sheet metal forming process of a cross member of automotive body.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.34 no.12
• /
• pp.1793-1803
• /
• 2010

The weighting method and goal programming require weighting factors or target values to obtain a Pareto optimal solution. However, it is difficult to define these parameters, and a Pareto solution is not guaranteed when the choice of the parameters is incorrect. Recently, the Mahalanobis Taguchi System (MTS) has been introduced to minimize the Mahalanobis distance (MD). However, the MTS method cannot obtain a Pareto optimal solution. We propose a function called the skewed Mahalanobis distance (SMD) to obtain a Pareto optimal solution while retaining the advantages of the MD. The SMD is a new distance scale that multiplies the skewed value of a design point by the MD. The weighting factors are automatically reflected when the SMD is calculated. The SMD always gives a unique Pareto optimal solution. To verify the efficiency of the SMD, we present two numerical examples and show that the SMD can obtain a unique Pareto optimal solution without any additional information.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2006.11a
• /
• pp.392-395
• /
• 2006

In this paper, Maharanobis distance was used to distinguish defective motors from good motors. Maharanobis distance was calculated from the noise data of good motors and the test motor that were measured in 1/3 octave hand from 25 Hz to 20 kHz frequency range. The suggested method was applied to the detection of defective air-conditioner motors.

• Journal of Korea Multimedia Society
• /
• v.14 no.6
• /
• pp.711-718
• /
• 2011

The k-Nearest Neighbor classifier that does not require a training phase is appropriate for a variable number of classes problem like face recognition, Recently distance metric learning methods that is trained with a given data set have reported the significant improvement of the kNN classifier. However, the performance of a distance metric learning method is variable for each application, In this paper, we focus on the face recognition and compare the performance of the state-of-the-art distance metric learning methods, Our experimental results on the public face databases demonstrate that the Mahalanobis distance metric based on PCA is still competitive with respect to both performance and time complexity in face recognition.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.17 no.2
• /
• pp.165-170
• /
• 2009

Pressure ripple, vibration and noise level are measured in each parts of the power steering system. MD(Mahalanobis Distance) is calculated by using MTS(Mahalanobis Taguchi System) with measured data, and noise sensitive components are selected. The components applied detail design parameters are made and data is measured. After that MD is calculated also. Mean value and SN ratio can be obtained from the MD. Effective noise reduction technique and dominant design parameters in hydraulic power steering system are introduced.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2024.05a
• /
• pp.764-767
• /
• 2024

웨어러블 디바이스에서는 사용자의 다양한 메타데이터를 수집할 수 있다. 그러나 이런 개인정보를 함유하고 있는 데이터를 수집하는 것은 사용자에게 개인정보침해 위협을 야기한다. 때문에 본 논문에서는 개인정보보호를 통한 웨어러블 디바이스 데이터활용방안으로 연합학습을 채택하였다. 다만 기존 연합학습에서도 해결해야할 문제점들이 있다. 우리는 그중에서도 데이터이질성(Data Heterogeneity) 문제해결을 위해 군집화(Clustering) 방법을 활용하였다. 또한 기존의 코사인유사도 기반 군집화에서 파라미터중요도가 반영되지 않는다는 문제점을 해결하고자 데이터수 기반 마하라노비스거리(Number of Samples Mahalanobis Distance) 군집화 방법을 제시하였다. 이를 통해 WESAD(Werable Stress Affect Detection)데이터에서 피실험자의 데이터 이질성이 존재하는 상황에서 기존 연합학습보다 학습 안정성 측면에서 좋음을 보여주었다.

• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.22 no.1
• /
• pp.22-28
• /
• 2013

Robotic GMA (Gas Metal Arc) welding process is one of widely acceptable metal joining process. The heat and mass inputs are coupled and transferred by the weld arc to the molten weld pool and by the molten metal that is being transferred to the weld pool. The amount and distribution of the input energy are basically controlled by the obvious and careful choices of welding process parameters in order to accomplish the optimal bead geometry and the desired quality of the weldment. To make effective use of automated and robotic GMA welding, it is imperative to predict online faults for bead geometry and welding quality with respect to welding parameters, applicable to all welding positions and covering a wide range of material thickness. MD (Mahalanobis Distance) technique was employed for investigating and modeling the GMA welding process and significance test techniques were applied for the interpretation of the experimental data. To successfully accomplish this objective, two sets of experiment were performed with different welding parameters; the welded samples from SM 490A steel flats. First, a set of weldments without any faults were generated in a number of repeated sessions in order to be used as references. The experimental results of current and voltage waveforms were used to predict the magnitude of bead geometry and welding quality, and to establish the relationships between weld process parameters and online welding faults. Statistical models developed from experimental results which can be used to quantify the welding quality with respect to process parameters in order to achieve the desired bead geometry based on weld quality criteria.

• Journal of KIISE:Computing Practices and Letters
• /
• v.9 no.3
• /
• pp.301-310
• /
• 2003

In this paper, we design and Implement the supervised classification systems for high resolution satellite images. The systems support various interfaces and statistical data of training samples so that we can select the m()st effective training data. In addition, the efficient extension of new classification algorithms and satellite image formats are applied easily through the modularized systems. The classifiers are considered the characteristics of spectral bands from the selected training data. They provide various supervised classification algorithms which include Parallelepiped, Minimum distance, Mahalanobis distance, Maximum likelihood and Fuzzy theory. We used IKONOS images for the input and verified the systems for the classification of high resolution satellite images.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.9C
• /
• pp.876-883
• /
• 2003

In this paper, we propose an algorithm for organ recognition in ultrasound images using log power spectrum. The main procedure of the algorithm consists of feature extraction and feature classification. In the feature extraction, as a translation invariant feature, log power spectrum is used for extracting the information on echo of the organs tissue from a preprocessed input image. In the feature classification, Mahalanobis distance is used as a measure of the similarity between the feature of an input image and the representative feature of each class. Experimental results for real ultrasound images show that the proposed algorithm yields the improvement of maximum 30% recognition rate than the recognition algorithm using power spectrum and Euclidean distance, and results in better recognition rate of 10-40% than the recognition algorithm using weighted quefrency complex cepstrum.

• Korean Journal of Remote Sensing
• /
• v.30 no.2
• /
• pp.207-218
• /
• 2014

The information on the land cover along stream corridor is important for stream restoration and maintenance activities. This study aims to review the different classification methods for mapping the status of stream corridors in Seom River using airborne RGB and CIR digital ortho imagery with a ground pixel resolution of 0.2m. The maximum likelihood classification, minimum distance classification, parallelepiped classification, mahalanobis distance classification algorithms were performed with regard to the improvement methods, the skewed data for training classifiers and filtering technique. From these results follows that, in aerial image classification, Maximum likelihood classification gave results the highest classification accuracy and the CIR image showed comparatively high precision.