• 지능정보연구
• /
• 제14권2호
• /
• pp.193-206
• /
• 2008

의료 진단 문제는 기정의된 특성치들로 표현되는 환자의 상태 데이터로부터 병의 유무를 판단하는 일종의 분류 문제로 간주할 수 있다. 본 연구는 혼용 유전자 알고리즘 기반의 분류방법을 도입함으로써 의료 진단 문제와 같은 다차원의 패턴 분류 문제를 해결할 수 있는 방안을 제안하고 있다. 일반적으로 분류 문제는 데이터 패턴에 존재하는 여러 클래스 간 구분경계를 생성하는 접근방법을 사용하는데, 이를 위해 본 연구에서는 일단의 영역 에이전트들을 도입하여 이들을 유전자 알고리즘 및 국소 적응조작을 혼용함으로써 데이터 패턴에 적응하도록 유도하고 있다. 일반적인 유전자 알고리즘의 진화단계를 거친 에이전트들에 적용되는 국소 적응조작은 영역 에이전트의 확장, 회피 및 재배치로 이루어지며, 각 에이전트의 적합도에 따라 이들 중 하나가 선택되어 해당 에이전트에 적용된다. 제안된 의료 진단용 분류 방법은 UCI 데이터베이스에 있는 잘 알려진 의료 데이터, 즉 간, 당뇨, 유방암 관련 진단 문제에 적용하여 검증하였다. 그 결과, 기존의 대표적인 분류기법인 최단거리이웃방법(the nearest neighbor), C4.5 알고리즘에 의한 의사 결정트리(decision tree) 및 신경망보다 우수한 진단 수행도를 나타내었다.

• 조명전기설비학회논문지
• /
• 제23권5호
• /
• pp.18-28
• /
• 2009

고정자 권선의 단락으로 인한 고장을 해석하고 진단 알고리즘의 효과적인 시험 평가를 위해 사용될 수 있는 인버터 구동 영구자석 동기전동기의 고장모델이 제시된다. 기존에 전동기의 해석과 제어에 많이 사용되는 dq 모델은 상전압 모델을 변환한 것으로 전동기 고정자의 권선 단락 시에는 더 이상 3상평형 조건이 성립하지 않기 때문에 인버터 극전압으로부터 전동기 입력 전압을 구하기가 쉽지 않아 고장모델의 해석을 위해서 사용하기 어렵다. 이를 해결하기 위해 전동기 3상 변수와 선전압 관계식을 이용한 전동기의 고장모델이 제안된다. 제안된 고장모델의 타당성을 입증하기 위해 시뮬레이션이 수행되며 내부 고정자의 권선 단락이 가능하도록 제작된 전동기와 DSP TMS320F28335를 이용한 제어 시스템에 의해 동일 고장 조건에서 비교 실험이 수행된다.

• 대한방사선기술학회지:방사선기술과학
• /
• 제23권1호
• /
• pp.39-47
• /
• 2000

In this study, we showed a comparison and analysis making use of DWI(diffusion weighted image) using early diagnosis of cerebral Infarction and with the classified T2 weighted image, FLAIR images signal intensity for brain infarction period. period of cerebral infarction after the condition of a disease by ischemic stroke. To compare 3 types of image, we performed polynomial warping and affined transform for image matching. Using proposed algorithm, calculated signal intensity difference between T2WI, DWI, FLAIR and DWI. The quantification values between hand made and calculated data are almost the same. We quantified the each period and performed pseudo color mapping by comparing signal intensity each other according to previously obtained hand made data, and compared the result of this paper according to obtained quantified data to that of doctors decision. The examined mean and standard deviation for each brain infarction stage are as follows ; the means and standard deviations of signal intensity difference between DWI and T2WI for each period are $197.7{\pm}6.9$ in hyperacute, $110.2{\pm}5.4$ in acute, and $67.8{\pm}7.2$ in subacute. And the means and standard deviations of signal intensity difference between DWI and FLAIR for each period are $199.8{\pm}7.5$ in hyperacute, $115.3{\pm}8.0$ in acute, and $70.9{\pm}5.8$ in subacute. We can quantificate and decide cerebral infarction period objectively. According to this study, DWI is very exact for early diagnosis. We classified the period of infarction occurrence to analyze the region of disease and normal region in DW, T2WI, FLAIR images.

• Journal of Advanced Marine Engineering and Technology
• /
• 제36권5호
• /
• pp.662-673
• /
• 2012

This study introduces a condition diagnosis technique for a turbine governor system. The governor system is an important control system to handle turbine speed in a nuclear power plant. The turbine governor system includes turbine valves and stop valves which have their own functions in the system. Because a turbine governor system is operated by high oil pressure, it is very difficult to maintain under stable operating conditions. Turbine valves supply oil pressure to the governor system for proper operation. Using the pressure variation of turbine and governor valves, operating conditions of the turbine governor control system are detected and identified. To achieve automatic detection of valve status, time-based and frequency-based analysis is employed. In this study, a new approach, wavelet decomposition, was used to extract specific features from the pressure signals of the governor and stop valves. The extracted features, which represent the operating conditions of the turbine governor system, include important information to control and diagnose the valves. After extracting the specific features, decision rules were used to classify the valve conditions. The rules were generated by a decision tree algorithm (a typical simple method for data-based rule generation). The results given by the wavelet-based analysis were compared to detection results using time- and frequency-based approaches. Compared with the several related studies, the wavelet transform-based analysis, the proposed in this study has the advantage of easier application without auxiliary features.

• Journal of Electrical Engineering and Technology
• /
• 제9권1호
• /
• pp.372-377
• /
• 2014

Being able to monitor the heart will allow the diagnosis of heart diseases for patients during daily activities, and the detection of burden on the heart during strenuous exercise. Furthermore, with the help of U-health technology, immediate medical action can be taken, in the case of abnormal symptoms of the heart in daily life. Therefore, it appears to be necessary to develop the corresponding technology to monitor the condition of the heart daily. In this study, a novel wearable smart system was proposed, to monitor the activity of the heart in daily life, and to further evaluate the rhythm of arrhythmia. The wearable system includes three modified bipolar conductive fiber electrodes in the chest part, which can resolve the reduction problem of the magnitude of the signal, by magnifying the signal and removing the noise, to obtain high affinity and validity for medical-type usage (<0.903%). The biological signal acquisition and data lines, and the signal processing engine and communication consist of a conductive ink, and the pic18 and ANT protocol nRF24AP2, respectively. The proposed algorithm was able to detect a strong ECG, signal and r-point passing over the noise. The confidence intervals were 96 %, which could satisfy the requirement to detect arrhythmia under the unconstrained conditions.

• 한국지능시스템학회논문지
• /
• 제11권9호
• /
• pp.833-836
• /
• 2001

공장 자동화 및 무인 자동화를 실현하기 위해 가장 기본적이며 중요한 기술은 제조공정에 대한 감시 기술이다. 특히 절삭공정에서 생산성을 향상시키기 위해서는 절삭 과정 중 드릴이나 앤드밀 등과 같은 공구의 마모상태가 실시간으로 감시되어야 한다. 본 논문은 드릴 공정에서 퍼지 논리를 도입하여 마모진단 시스템을 구성하였다. 실시간 마모진단을 위해서 절삭력과 절삭력의 변화량을 퍼지 입력 변수로 하여 컴퓨터를 이용한 드릴의 마모상태를 판단하는 알고리즘을 제안하였다. 제안된 퍼지 마모진단 시스템을 평가하기 위하여 퍼지 마모량과 드릴의 실제 마모량을 측정하여 그 결과를 비교하였다.

• Nuclear Engineering and Technology
• /
• 제52권10호
• /
• pp.2262-2273
• /
• 2020

Digital twin technology can provide significant value for the prognostics and health management (PHM) of critical plant components by improving insight into system design and operating conditions. Digital twinning of systems can be utilized for anomaly detection, diagnosis and the estimation of the system's remaining useful life in order to optimize operations and maintenance processes in a nuclear plant. In this regard, a conceptual framework for the application of digital twin technology for the prognosis of Control Element Drive Mechanism (CEDM), and a data-driven approach to anomaly detection using coil current profile are presented in this study. Health management of plant components can capitalize on the data and signals that are already recorded as part of the monitored parameters of the plant's instrumentation and control systems. This work is focused on the development of machine learning algorithm and workflow for the analysis of the CEDM using the recorded coil current data. The workflow involves features extraction from the coil-current profile and consequently performing both clustering and classification algorithms. This approach provides an opportunity for health monitoring in support of condition-based predictive maintenance optimization and in the development of the CEDM digital twin model for improved plant safety and availability.

• 한국융합학회논문지
• /
• 제3권4호
• /
• pp.15-21
• /
• 2012

초음파 영상 진단 장치에서 획득한 데이터로부터 진단 객체를 추출하기 위한 영상 분할은 질병의 효과적인 진단을 위하여 필수적인 전처리 과정으로 인식되고 있으며, 지금까지 많은 분할 기법들이 연구되고 있다. 본 연구에서는 혈관 초음파 영상의 다양한 응용 및 진단법 개발을 위하여 기초 전처리과정으로서 graph cut 알고리즘에 의한 상호적인 영상분할법을 제시한다. 일반영상 및 혈관 초음파 영상에 대하여 전경(foreground)과 배경(background)의 제약조건을 주고 영상분할 처리하여, 원하는 object에 대한 분할 결과를 얻었다. 향후, 이러한 일련의 처리 과정이 실시간으로 처리되면 새로운 초음파 진단법으로 발전시켜 나갈 수 있을 것으로 사료된다.

• 전기전자학회논문지
• /
• 제25권1호
• /
• pp.212-217
• /
• 2021

Premature contraction arrhythmia is the most common disease among arrhythmia and it may cause serious situations such as ventricular fibrillation and ventricular tachycardia. Most of arrhythmia clasification methods have been developed with the primary objective of the high detection performance without taking into account the computational complexity. Also, personalized difference of ECG signal exist, performance degradation occurs because of carrying out diagnosis by general classification rule. Therefore it is necessary to design efficient method that classifies arrhythmia by analyzing the persons's physical condition and decreases computational cost by accurately detecting minimal feature point based on only QRS features. We propose method for personalized specific classification of premature contraction arrhythmia based on QRS features in smart healthcare environments. For this purpose, we detected R wave through the preprocessing method and SOM and selected abnormal signal sets.. Also, we developed algorithm to classify premature contraction arrhythmia using QRS pattern, RR interval, threshold for amplitude of R wave. The performance of R wave detection, Premature ventricular contraction classification is evaluated by using of MIT-BIH arrhythmia database that included over 30 PVC(Premature Ventricular Contraction) and PAC(Premature Atrial Contraction). The achieved scores indicate the average of 98.24% in R wave detection and the rate of 97.31% in Premature ventricular contraction classification.

• 한국기계가공학회지
• /
• 제21권3호
• /
• pp.77-85
• /
• 2022

In this study, to diagnose the grinding state of a micro drill bit, a sensor attachment location was selected through random vibration analysis of the grinding unit of the micro drill-bit grinding system. In addition, the vibration data generated during the drill bit grinding were collected from the grinding unit for the grinding wheels under the steady and worn conditions, and data feature extraction and dimension reduction were performed. The wear of the micro-drill-bit grinding wheel was diagnosed by applying KNN, a machine-learning algorithm. The classification model showed excellent performance, with an accuracy of 99.2%. The precision, recall and f1-score were higher than 99% in both the steady and wear conditions.