• 해양환경안전학회지
• /
• 제27권7호
• /
• pp.1088-1097
• /
• 2021

기계 장비의 진동 데이터는 필연적으로 노이즈를 포함하고 있다. 이러한 노이즈는 기계 장비의 유지보수를 진행하는데 악영향을 끼친다. 그에 따라 데이터의 노이즈를 얼마나 효과적으로 제거해주냐에 따라 학습 모델의 성능을 좌우한다. 본 논문에서는 시계열 데이터를 전처리 함에 있어 특성추출 과정을 포함하지 않는 Denoising Auto Encoder 기법을 활용하여 데이터의 노이즈를 제거했다. 또한 기계 신호 처리에 널리 사용되는 Wavelet Transform과 성능 비교를 진행했다. 성능비교는 고장 탐지율을 계산하여 진행했으며 보다 정확한 비교를 위해 분류 성능 평가기준 중 하나인 F-1 Score를 계산하여 성능 비교를 진행했다. 고장을 탐지하는 과정에서는 One-Class SVM 기법을 활용하여 고장 데이터를 탐지했다. 성능 비교 결과 고장 진단율과 오차율 측면에서 Denoising Auto Encoder 기법이 Wavelet Transform 기법에 비해 보다 좋은 성능을 나타냈다.

• 대한후두음성언어의학회지
• /
• 제14권1호
• /
• pp.16-25
• /
• 2003

후두암은 유병율이 높지만 조기에 발견하면 90% 이상의 치유율과 발성기능의 보존이 가능하며 현재 음성분석을 이용한 진단법이 시도되고 있으나 정립된 선별검사법은 없는 실정이다. 성문전도검사(electroglottography, EGG)는 성대의 진동양상을 알 수 있는 비침습적 검사로서 발성과 음성합성의 연구에 많이 사용되고 있다. 본 연구는 EGG에서 관찰되는 파라미터들을 다층 퍼셉트론(multilayer perceptron)구조의 신경회로망(artificial neural network)으로 감별하는 기법을 이용하여 후두암 감별법에 대한 연구로서 부산대학교병원을 내원한 후두암 환자 10명과 양성후두질환 26명을 대상으로 새로 고안한 Electroglottograph(v1.0)를 이용하여 검사하고 이의 임상적 유용성을 평가하였다. EGG 파라미터인 closed quotient(CQ), speed quotient(SQ), speed index(SI), fundamental frequency(F0), Jitter, Shimmer 등은 MATLAB 6.5 (Mathwork, Inc.)로 작성한 분석 프로그램을 이용하여 추출하였다. 각 환자에서 추출된 EGG 파라미터들을 다층 퍼셉트론 구조의 신경회로망으로 감별하였다. CQ는 각 질환군 간에 유의한 차이가 없었지만 SQ, SI, Jitter, Shimmer 등은 성대질환의 특성에 따라 유의한 차이를 보였다. 신경회로망에서 감별한 결과 CQ를 제외한 SQ, SI, Jitter, Shimmer 등에서 71.3-90%의 후두암의 감별율을 보였다. 또한 SQ, SI, Jitter, Shimmer를 3개씩 조합한 실험에서는 SQ-Jitter-Shimmer와 SQ-SI-Shimmer의 후두암의 감별율이 93%로 가장 높았고, SQ-SI-Jitter 90.9%, SI-Jitter-Shimmer 88.6%로 전체적으로 85% 이상의 높은 감별율을 나타내었다. 이러한 결과는 EGG검사와 신경회로망을 이용한 양성과 악성 후두질환의 감별이 가능함을 시사한다. 향후 성대 질환의 병태생리를 대변할 수 있는 파라미터가 추가로 개발되고 분류 알고리듬이 개선된다면 EGG를 이용한 성대질환의 감별 진단이 보다 정확해질 것으로 사료되었다.

• 한국전문물리치료학회지
• /
• 제4권2호
• /
• pp.18-35
• /
• 1997

The purpose of this study were to determine causes of low back pain in bus drivers who usually work in prolonged sitting position and to find the relationship of back pain with general characteristics and work environmental characteristics, stress symptoms, to com맹re lordosis angle of habitual driving posture and resting posture in low back pain group and non-low back pain group, and to use this result as a basis for improvement of work environment and comprehensive rehabilitative management of low back pain. Ninety-eight bus drivers were selected from the membership of an urban transit union in Seoul. These informations were collected from May 1, 1997 to May 25, 1997 by means of structured questionnaires and X-ray findings. These data were analyzed by $x^2$ test, t-test, logistic regression using SAS. The major results were as follows: 1. Of the respondents, 66.3% of bus drivers were found to be experiencing back pain. 2. Of the respondents of low back pain groups, 78.5% reported that major cause of low back pain was due to prolonged sitting. 3. The group with experiences of frequent or continuous vibration had more low back pain (p<0.05). There were no significant differences among other work environmental factors. 4. Average stress score was significantly higher in the group with low back pain than in the group without low back pain (p<0.05). 5. The results of the logistic regression analyses were statistically significant vibration from seat and stress score among the risk factors (p<0.05). 6. Averaged lordosis angle of habitual driving posture was $7.6{\pm}14.9$ degrees in low back pain group, $16.5{\pm}8.7$ degrees in non-low back pain and averaged lordosis angle of resting sitting posture was $10.8{\pm}13.7$ degrees in low back pain, $18.9{\pm}9.6$ degrees in non-low back pain group but the difference in mean lordosis angles of two groups was not statistically significant. The results of this study indicate that professional drivers developed chronicity of low back pain due to unfit seat, poor habitual posture and stress from vibration or other risk factors. Therefore, there is need to improve work environment, i.e. enough resting, to set a seat to support lumbar spine properly and to provide comprehensive rehabilitation program including early diagnosis, proper treatment and education for self help management.

• 대한한의학회지
• /
• 제32권6호
• /
• pp.10-17
• /
• 2011

Background: Vibration response imaging (VRI) is a new technology that records energy generated by airflow during the respiration cycle. Analysis of lung sound using VRI may overcome the limitations of auscultation. Objectives: To set a VRI standard for healthy Koreans, we conducted a clinical assessment to evaluate breath sound images and quantification in healthy subjects and compared the findings with reported breath sound characteristics. Methods: Recordings were performed using the VRIxp. Eighty subjects took a deep breath four times during a 12-second interval while sitting upright. The quantitative aspect was analyzed using the VRI quantitative lung data (QLD) for total left lung, total right lung and for six lung regions: left upper lung (LUL), left middle lung (LML), left lower lung (LLL), right upper lung (RUL), right middle lung (RML), right lower lung (RLL). The qualitative aspect was provided through image assessments by three reviewers. Results: In all regions the left lung had significantly higher QLD than the right lung (P<0.005, paired t-test). The inter-rater agreement was 0.78. 84% of the images were found normal by the final assessment. Among the 16% (n=13) of images with abnormal final assessment, the most common flawed features were dynamic image (77%, n=10) and maximum energy frame (MEF) shape (77%, n=10). No significant differences were found between males and females for QLD but there were significant differences in qualitative aspects including dynamic images, MEF shape, and missing LLL. Conclusion: The characteristics of healthy Koreans are similar to those of Western subjects reported previously. VRI is easy to use and objective, and so is helpful to diagnose patients with respiratory diseases and to monitor the progress of diseases after medical treatments.

• 대한방사선기술학회지:방사선기술과학
• /
• 제31권1호
• /
• pp.61-69
• /
• 2008

유방초음파 검사는 유방암 검사에 있어 유방촬영술에 비하여 많은 장점이 있으나, 미세석회화 발견에는 적합하지 않은 단점이 있다. 이에 유방초음파 검사에서 기존의 7.5 MHz 선형 탐촉자를 사용하여 매질의 공명현상을 이용한 유방조직 석회화 병변을 관측할 수 있는 방법을 연구하였다. 먼저 gelatin과 돼지 젖가슴살을 이용하여 유방조직 팬텀을 제작하였으며, 외부 진동을 변화시켜 가며 석회화 병변을 관측하였다. 유방조직 팬텀안에 주입된 석회화는 주변 조직과 다른 공명을 일으키면서 외부진동에 따라 음향공명의 정도가 파워도플러의 ROI 영역 내의 색상의 밝기와 영역의 차이로 나타내었다. 낮은 주파수 영역에는 음향공명이 거의 나타나지 않았으며, 약 $300{\sim}400\;Hz$ 사이에서 일정한 플래토우 영역을 나타내었으며, 이후 주파수가 증가함에 따라 색상이 사라짐을 확인하였다.

• 한국정밀공학회지
• /
• 제16권12호
• /
• pp.119-125
• /
• 1999

This paper gives an overview of the various approaches to designing statistical pattern recognition scheme based on Bayes discrimination rule and the artificial neural networks for rotating machine condition classification. Concerning to Bayes discrimination rule, this paper contains the linear discrimination rule applied to classification into several multivariate normal distributions with common covariance matrices, the quadratic discrimination rule under different covariance matrices. Also we discribes k-nearest neighbor method to directly estimate a posterior probability of each class. Five features are extracted in time domain vibration signals. Employing these five features, statistical pattern classifier and neural networks have been established to detect defects on rotating machine. Four different cases of rotation machine were observed. The effects of k number and neural networks structures on monitoring performance have also been investigated. For the comparison of diagnosis performance of these two method, their recognition success rates are calculated form the test data. The result of experiment which classifies the rotating machine conditions using each method presents that the neural networks shows the highest recognition rate.

• 한국해양공학회지
• /
• 제25권4호
• /
• pp.59-65
• /
• 2011

In structural health monitoring (SHM) using electro-mechanical impedance signatures, it is a critical issue for extremely large structures to extract the best damage diagnosis results, while minimizing unknown environmental effects, including temperature, humidity, and acoustic vibration. If the impedance signatures fluctuate because of these factors, these fluctuations should be eliminated because they might hide the characteristics of the host structural damages. This paper presents a long-term SHM technique under an unknown noisy environment for tidal current power plant structures. The obtained impedance signatures contained significant variations during the measurements, especially in the audio frequency range. To eliminate these variations, a continuous principal component analysis was applied, and the results were compared with the conventional approach using the RMSD (Root Mean Square Deviation) and CC (Cross-correlation Coefficient) damage indices. Finally, it was found that this approach could be effectively used for long-term SHM in noisy environments.

• 한국융합학회논문지
• /
• 제11권11호
• /
• pp.47-53
• /
• 2020

최근 모니터링 및 예측 시스템을 이용하여 사전에 결함을 발견하고 이를 경고하는 시스템이 활발히 연구되고 있다. 차량 안전 관리에 있어서도 예측 결함 분석 기술을 적용하여 자동차 휠 베어링의 고장 유무 및 고장 유형을 조기에 경고하는 시스템이 필요하다. 본 논문에서는 휠 베어링과 결합 된 센서 모듈과 각 센서 모듈에서 차량 가속 정보 및 진동 정보를 수집, 저장 및 분석하는 진단 시스템을 제시하였다. 제안된 센서 모듈은 저비용으로 차량의 휠 베어링 상태를 모니터링하며, 이렇게 수집된 데이터를 활용하여 진단 및 고장 예측 기능을 수행하는 방안을 연구하였다. 개발된 센서 모듈과 예측 분석 시스템은 가진 테스트 장비 및 실제 차량을 이용하여 테스트하고 그 유효성을 평가하였다.

• 대한후두음성언어의학회지
• /
• 제19권1호
• /
• pp.58-61
• /
• 2008

A 66-year-old woman described a 6-month history of hoarseness after upper respiratory infection. She was a house wife and referred from an outside clinic under a diagnosis of vocal fold nodules. Strobovideolaryngoscopy revealed bilateral vocal fold lesions, and decreased mucosal wave of both vocal folds. She was brought to the operating room for microsuspension laryngoscopy. Under general anesthesia, dual intracordal cysts on left vocal fold were completely resected with microflap technique. The lesion on the right vocal fold turned out to be a reactive fibrous mass, which was also resected. Dual intracordal cysts were confirmed histopathologically. The one was an epidermoid cyst lined with squamous epithelium, and the other was a mucus retention cyst lined with cuboidal epithelium. Postoperative voice was acceptable by the patient and the mucosal vibration has much improved after the surgery.

• Smart Structures and Systems
• /
• 제13권5호
• /
• pp.869-890
• /
• 2014

In this study, a damage detection approach using train-induced vibration response of the bridge is proposed, utilizing only direct structural analysis by means of introducing soft computing methods. In this approach, the possible damage patterns of the bridge are assumed according to theoretical and empirical considerations at first. Then, the running train-induced dynamic response of the bridge under a certain damage pattern is calculated employing a developed train-bridge interaction analysis program. When the calculated result is most identical to the recorded response, this damage pattern will be the solution. However, owing to the huge number of possible damage patterns, it is extremely time-consuming to calculate the bridge responses of all the cases and thus difficult to identify the exact solution quickly. Therefore, the soft computing methods are introduced to quickly solve the problem in this approach. The basic concept and process of the proposed approach are presented in this paper, and its feasibility is numerically investigated using two different train models and a simple girder bridge model.