• International Journal of Aeronautical and Space Sciences
• /
• v.9 no.1
• /
• pp.100-110
• /
• 2008

It is not easy to monitor and identify all engine faults and conditions using conventional fault detection approaches like the GPA (Gas Path Analysis) method due to the nature and complexity of the faults. This study therefore focuses on a model based diagnostic method using Neural Network algorithms proposed for fault detection on a turbo shaft engine (PW 206C) selected as the power plant for a tilt rotor type unmanned aerial vehicle (Smart UAV). The model based diagnosis should be performed by a precise performance model. However component maps for the performance model were not provided by the engine manufacturer. Therefore they were generated by a new component map generation method, namely hybrid method using system identification and genetic algorithms that identifies inversely component characteristics from limited performance deck data provided by the engine manufacturer. Performance simulations at different operating conditions were performed on the PW206C turbo shaft engine using SIMULINK. In order to train the proposed BPNN (Back Propagation Neural Network), performance data sets obtained from performance analysis results using various implanted component degradations were used. The trained NN system could reasonably detect the faulted components including the fault pattern and quantity of the study engine at various operating conditions.

• Journal of Korea Multimedia Society
• /
• v.17 no.3
• /
• pp.277-286
• /
• 2014

In this paper, we propose a texture feature-based language identification by fusion of Gabor, MDLC (multi-lag directional local correlation), and co-occurrence features. In the proposed method, for a test image, Gabor magnitude images are first obtained by Gabor transform followed by magnitude operator. Moments for the Gabor magniude images are then computed and vectorized. MDLC images are then obtained by MDLC operator and their moments are computed and vectorized. GLCM (gray-level co-occurrence matrix) is next calculated from the test image and co-occurrence features are computed using the GLCM, and the features are also vectorized. The three vectors of the Gabor, MDLC, and co-occurrence features are fused into a feature vector. In classification, the WPCA (whitened principal component analysis) classifier, which is usually adopted in the face identification, searches the training feature vector most similar to the test feature vector. We evaluate the performance of our method by examining averaged identification rates for a test document image DB obtained by scanning of documents with 15 languages. Experimental results show that the proposed method yields excellent language identification with rather low feature dimension for the test DB.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.26 no.7
• /
• pp.844-850
• /
• 2016

This study pertains to the extraction of the road noise component of signals from a vehicle's interior noise via the traditional frequency domain and time domain system identification methods. For road noise extraction based on the frequency domain system identification method, the appropriate matrix inversion strategy is investigated and causal and non-causal impulse response filters are compared. Furthermore, appropriate data lengths for the frequency domain system identification method are investigated. In addition to the traditional road noise extraction methods based on frequency domain system identification, a new approach to extract road noise via the time domain system identification method based on a parametric input-output model is proposed and investigated in the present study. In this approach, instead of constructing a higher order model for the full-band road noise, input and output signals are processed in the subband domain and lower order parametric models optimal to each subband are determined. These parametric models are used to extract road noises in each subband; the full band road noise is then reconstructed from the subband road noises. This study shows that both the methods in the frequency domain and the time domain successfully extract the road noise from the vehicle's interior noise.

• Korean Journal of Breeding Science
• /
• v.42 no.1
• /
• pp.50-55
• /
• 2010

The R-r:standard (R-r:std) allele of maize R gene complex consists of S subcomplex and P component; the S subcomplex regulates anthocyanin pigmentation of seed aleurone layer, and the P component confers pigmentation of the other plant parts. The S subcomplex contains two functional genes, S1 and S2 components. In the presence of Pl gene some alleles of R gene induce anthocyanin pigmentation of pericarp. In the present study, the effects of different R alleles on the anthocyanin pigmentation of pericarp in the presence of Pl gene were analyzed in order to identify the R gene component responsible for pericarp pigmentation. The results show that R-ch and r-ch alleles condition similar degrees of pericarp pigmentation, and that R-r:Ecuador (R-r:Ec) conditions stronger pigmentation. The r-ch allele, which is inferred that its S subcomplex has lost function but the P component is normal, induces pericarp pigmentation in the presence of Pl gene. On the contrary, the R-g:g1111 allele, derived from R-r:Ec and inferred that its S subcomplex functions normal but the P component has lost its function, did not induce pericarp pigmentation in the presence of Pl gene. Moreover, PCR analysis of genomic DNA's of R-ch and r-ch indicate that R-ch maintains both P and S1 components, whereas r-ch lacks for the S1 component. Taken together, The results suggest that the P components of R alleles inducing pericarp pigmentation in the presence of Pl gene are responsible for pericarp pigmentation.

• Journal of Korea Multimedia Society
• /
• v.23 no.9
• /
• pp.1155-1163
• /
• 2020

Deep learning approach is emerging as a new way to improve the accuracy of tree species identification using bark image. However, the approach has not been studied enough because it is confronted with the problem of acquiring a large volume of bark image dataset. This study solved this problem by utilizing a pretrained off-the-shelf DCNN model. It compares the discrimination power of bark features extracted by each DCNN model. Then it extracts the features by using a selected DCNN model and feeds them to a multi-layer perceptron (MLP). We found out that the ResNet50 model is effective in extracting bark features and the MLP could be trained well with the features reduced by the principal component analysis. The proposed approach gives accuracy of 99.1% and 98.4% for BarkTex and Trunk12 datasets respectively.

• The Proceedings of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.5 no.3
• /
• pp.54-63
• /
• 1991

This paper describes an effort to develope a microcomputer-based parameter measurement system for a brushless DC motor. Back EMF equation is derived from back EMF waveform of a brushless DC motor. To minimize error the due to the ripple component in the measured armature current, digital averaging filter is employed. The whole identification process of signal generation, measurement parameter determination is fully automated. A new identification algorithm for the brushless DC motor parameters is developed. New parameter correction method is proposed using the deadzone current and the time to reach the peak current. In the proposed correction method, the measured current is in excellent agreement with the estimated current.

• Journal of KSNVE
• /
• v.9 no.4
• /
• pp.761-768
• /
• 1999

This paper presents an NVH test of Automotive power steering system performed at a half-car Test-rig. The test was done for neutral and full turn(or relief) conditions in steering wheel at a fixed rpm first, then followed by the same conditions for the rpm run-up. The sound intensity measurement verified the results from the frequency and order analysis, especially about the identification of major noise sources and their dominant frequencies. The results from thie study can be utilized in the system noise tuning when a new steering component is installed. In particular, the noise and vibration reduction at the relief condition will be accomplished through the knowledge obtained from this study and from the on-going research on the hose tuning techniques usign silencers and tuning cable inserted in the pressure hose.

• Proceedings of the Korea Concrete Institute Conference
• /
• 2000.10a
• /
• pp.363-368
• /
• 2000

This paper is an introduction of pessimum program for the identification of alkali-silica reaction of alkali-aggregate reaction which is known as one of a major factor of concrete deterioration. A series of gel-pat testing program was undertaken to observe the reactivity of potentially alkali-silica reactive concrete aggregates which were found to be reactive by previous petrographic examination (ASTM C 295). And then a pessimum program was performed in accordance with mortar-bar test method (ASTM C 227) with different percentage of those reactive components included in the fine aggregate source to determine the pessimum quantity. Chert and quartzite were found to be major components of reactive mineral/rock, and the pessimum condition for chert was about 3%, even though the test was performed with up to 25% of the component. In the case of quartzite, however, the mortar-bar expansion appeared to be directly proportional to the amount of quartzite sample with increasing tested quantity up to 35%. Both of the expansion results were well 3 and 6 month specified maximum limitation of 0.05% and of 0.1% respectively.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.6 no.2
• /
• pp.346-355
• /
• 2002

This study will offer multimodal recognition instead of an existing monomodal bioinfomatics by using face, lips, to improve the accuracy of recognition. Each bioinfomatics vector can be found by the following ways. For a face, the feature is calculated by principal component analysis with wavelet multiresolution. For a lip, a filter is used to find out an equation to calculate the edges of the lips first. Then by using a thinning image and least square method, an equation factor can be drawn. A voice recognition is found with MFCC by using mel frequency. We＇ve sorted backpropagation neural network and experimented with the inputs used above. Based on the experimental results we discuss the advantage and efficiency.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.3 no.1
• /
• pp.13-25
• /
• 2000

A simulation model is developed to analyze the large-structure driving system and its integrated behavior in the whole weapon system. It models every component in the driving system such as mechanical and electrical characteristics, and it is programmed by simulation language in a way which strongly reflects the system's real time dynamics and reduces computation time as well. A useful parameter identification method is proposed, and it is tuned on the given physical system. The model is validated through comparing to real test, and it is applied to analysis and prediction of integrated system functions relating to the fire control system.