• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.15 no.1
• /
• pp.69-75
• /
• 2022

Due to the increase in the demand for anomaly detection according to the ageing of power generation facilities, the need for developing an anomaly detection method that can provide high-reliability turbine blade anomaly detection performance has been continuously raised. Additionally, the false detection results caused by a human error accelerates the increase of the need. In this paper, we propose an anomaly detection technique for turbine blades in power plants using deep neural networks. Experimental results prove that the proposed technique achieves stable anomaly detection performance while minimizing human factor intervention.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.2
• /
• pp.494-510
• /
• 2024

Internet users are exposed to sophisticated cyberattacks that intrusion detection systems have difficulty detecting. Therefore, research is increasing on intrusion detection methods that use artificial intelligence technology for detecting novel cyberattacks. Unsupervised learning-based methods are being researched that learn only from normal data and detect abnormal behaviors by finding patterns. This study developed an anomaly-detection method based on unsupervised machines and deep learning for a network intrusion detection system (NIDS). We present a hybrid anomaly detection approach based on unsupervised learning techniques using the autoencoder (AE), Isolation Forest (IF), and Local Outlier Factor (LOF) algorithms. An oversampling approach that increased the detection rate was also examined. A hybrid approach that combined deep learning algorithms and traditional machine learning algorithms was highly effective in setting the thresholds for anomalies without subjective human judgment. It achieved precision and recall rates respectively of 88.2% and 92.8% when combining two AEs, IF, and LOF while using an oversampling approach to learn more unknown normal data improved the detection accuracy. This approach achieved precision and recall rates respectively of 88.2% and 94.6%, further improving the detection accuracy compared with the hybrid method. Therefore, in NIDS the proposed approach provides high reliability for detecting cyberattacks.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2011.06a
• /
• pp.845-846
• /
• 2011

• Transactions of the Korean Society of Automotive Engineers
• /
• v.13 no.4
• /
• pp.74-80
• /
• 2005

This paper presents a misfire monitoring method by using the weight factor. According to OBD II(On-Board Diagnostics) regulations of the CARB (California Air Resources Board), an ECU (Electronic Control Unit) should detect misfires which occur in the internal combustion engine. A misfire is 1311owe4 by post-oscillations for short duration. Sometimes, the amplitude of oscillations may be as high as misfire and can be falsely detected as another misfire. To prevent this, the software designers do not attempt to detect another misfire for this short duration, during which the post oscillations exist. Because of this, ECU does not detect all the misfires and hence, the unstable state of automobile cannot be detected. If this happens for a long time, automobile may get damaged. To solve these problems, this paper suggests a new algorithm to detect misfire by using weighting factor Weighting factor is a concept to distinguish the misfire with the post oscillation and to improve the detection rate. This value of weighting factor is used for counting the misfire. This paper also shows the result of experiment done on a automobile using this software. The software is implemented using ASCET-SD which is preferred in the design of engine control. This paper's result show the possibility of improving the misfire detection by implementing this algorithm.

• The Korean Journal of Mycology
• /
• v.41 no.1
• /
• pp.52-55
• /
• 2013

Phoma glomerata (Corda) Wollenw. & Hochapfel is a pathogenic fungus causing spot diseases of plant leaves and fruits. This fungus is important in plant quarantine of seedlings and fruits in Korea. The aim of this study was to develop a sensitive and effective diagnostic method for P. glomerata detection in imported plants. The fungal species-specific PCR primers were designed based on the nucleotide sequences of the translation elongation factor 1 alpha gene and their specificity and sensitivity were tested. The designed primers named as PhoGlo-F and PhoGlo-R amplified specifically a 170 bp sized DNA band of the target gene from the genomic DNA of P. glomerata. No amplicon was produced from genomic DNAs of 16 other Phoma spp. and reference fungal species tested. Moreover, PhoGlo-F/PhoGlo-R primers successfully worked with real-time PCR technique. The detection limit of DNA content by conventional and real-time PCR were 10 pg and 1pg of the genomic DNA of P. glomerata, respectively. We believed that the developed makers would be very useful for P. glomerata detection.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.13 no.4
• /
• pp.39-46
• /
• 2013

Cooperative spectrum sensing can improve sensing reliability, compared with single node spectrum sensing. In addition, Eigenvalue-based spectrum sensing has also drawn a great attention due to its performance improvement over the energy detection method in which the more smoothing factor, the better performance is achieved. However, the more smoothing factor in Eignevalue-based spectrum sensing requires the more sensing time. Furthermore, more reporting time in cooperative sensing will be required as the number of nodes increases. Subsequently, we in this paper propose an Eigenvalue and superposition-based spectrum sensing where the reporting time is utilized so as to increase the number of smoothing factors for autocorrelation calculation. Simulation result demonstrates that the proposed scheme has better detection probability in both local as well as global detection while requiring less sensing time as compared with conventional Eigenvalue-based detection scheme.

• Smart Structures and Systems
• /
• v.14 no.3
• /
• pp.419-444
• /
• 2014

Within this paper a new approach for early damage detection in rotor blades of wind energy converters is presented, which is shown to have a more sensitive reaction to damage than eigenfrequency-based methods. The new approach is based on the extension of Gasch's proportionality method, according to which maximum oscillation velocity and maximum stress are proportional by a factor, which describes the dynamic behavior of the structure. A change in the proportionality factor can be used as damage indicator. In addition, a novel deflection sensor was developed, which was specifically designed for use in wind turbine rotor blades. This deflection sensor was used during the experimental tests conducted for the measurement of the blade deflection. The method was applied on numerical models for different damage cases and damage extents. Additionally, the method and the sensing concept were applied on a real 50.8 m blade during a fatigue test in the edgewise direction. During the test, a damage of 1.5 m length was induced on the upper trailing edge bondline. Both the initial damage and the increase of its length were successfully detected by the decrease of the proportionality factor. This decrease coincided significantly with the decrease of the factor calculated from the numerical analyses.

• KSBB Journal
• /
• v.28 no.4
• /
• pp.254-259
• /
• 2013

Herein we present a diagnostic paper chip that can quantitatively detect albumin without external electronic reader and dispensing apparatus. We fabricated a diagnostic paper chip device by printing wax barrier on the paper and wicking it with citrate buffer and tetrabromophenol blue to detect albumin in sample solution. The paper chip is so simple that we dropped a sample solution at sample pad and measure the ratio of two travel distances of the sample solvent and albumin under the name of retention factor. Our result confirmed that the retention factor was constant in the samples with same concentration of albumin and useful determinant for the measurement of albumin concentration. The paper chip is affordable and equipment-free, and close to ideal point-of-care test in accordance with the assured criteria, outlined by the World Health Organization. We assume that this diagnostic paper chip will expand the concept of colorimetric determination and provide a inexpensive diagnostic method to aging society and developing country.

• Genomics & Informatics
• /
• v.5 no.4
• /
• pp.143-151
• /
• 2007

To understand the mechanism of transcriptional regulation, it is essential to detect promoters and regulatory elements. Various kinds of methods have been introduced to improve the prediction accuracy of regulatory elements. Since there are few experimentally validated regulatory elements, previous studies have used criteria based solely on the level of scores over background sequences. However, selecting the detection criteria for different prediction methods is not feasible. Here, we studied the calibration of thresholds to improve regulatory element prediction. We predicted a regulatory element using MATCH, which is a powerful tool for transcription factor binding site (TFBS) detection. To increase the prediction accuracy, we used a regulatory potential (RP) score measuring the similarity of patterns in alignments to those in known regulatory regions. Next, we calibrated the thresholds to find relevant scores, increasing the true positives while decreasing possible false positives. By applying various thresholds, we compared predicted regulatory elements with validated regulatory elements from the Open Regulatory Annotation (ORegAnno) database. The predicted regulators by the selected threshold were validated through enrichment analysis of muscle-specific gene sets from the Tissue-Specific Transcripts and Genes (T-STAG) database. We found 14 known muscle-specific regulators with a less than a 5% false discovery rate (FDR) in a single TFBS analysis, as well as known transcription factor combinations in our combinatorial TFBS analysis.

• Journal of the military operations research society of Korea
• /
• v.11 no.2
• /
• pp.6-14
• /
• 1985

The layered multi-barrier defense situation against penetrating enemy threat is analytically modeled towards minimizing the penetration probability. Each layer is characterized by probability of detection and probability of kill given detection. The two capabilities are assumed independent. Detection in a layer, however, affects detection performance in subsequent layers. The following three models were formulated and investigated: (1) 'Model A' permits increase of detection performance in only the next barrier, (2) 'Model B' permits the increase in all subsequent barriers linearly, and (3) 'Model C' expresses the increase in an asymptotic exponential way. The best and the worst barrier combinations are determined through model exercise and model performances are compared through sensitivity analysis for the 'intensification factor.'