• Smart Structures and Systems
• /
• v.29 no.6
• /
• pp.757-766
• /
• 2022

To train deep learning algorithms, a sufficient number of data are required. However, in most engineering systems, the acquisition of fault data is difficult or sometimes not feasible, while normal data are secured. The dearth of data is one of the major challenges to developing deep learning models, and fault diagnosis in particular cannot be made in the absence of fault data. With this context, this paper proposes an anomaly detection methodology for rotating machines using only normal data with self-labeling. Since only normal data are used for anomaly detection, a self-labeling method is used to generate a new labeled dataset. The overall procedure includes the following three steps: (1) transformation of normal data to self-labeled data based on a pretext task, (2) training the convolutional neural networks (CNN), and (3) anomaly detection using defined anomaly score based on the softmax output of the trained CNN. The softmax value of the abnormal sample shows different behavior from the normal softmax values. To verify the proposed method, four case studies were conducted, on the Case Western Reserve University (CWRU) bearing dataset, IEEE PHM 2012 data challenge dataset, PHMAP 2021 data challenge dataset, and laboratory bearing testbed; and the results were compared to those of existing machine learning and deep learning methods. The results showed that the proposed algorithm could detect faults in the bearing testbed and compressor with over 99.7% accuracy. In particular, it was possible to detect not only bearing faults but also structural faults such as unbalance and belt looseness with very high accuracy. Compared with the existing GAN, the autoencoder-based anomaly detection algorithm, the proposed method showed high anomaly detection performance.

• Journal of Ship and Ocean Technology
• /
• v.6 no.4
• /
• pp.1-12
• /
• 2002

A control system for an automated line heating process is developed by use of object-oriented methodology. The main function of the control system is to provide real-time heating information to technicians or automated machines. The information includes heating location, torch speed, heating order, and others. The system development is achieved by following the five steps in the object-oriented procedure. First, requirements are specified and corresponding objects are determined. Then, the analysis, design, and implementation of the proposed system are sequentially carried out. The system consists of six subsystems, or modules. These are (1) the inference module with an artificial neural network algorithm, (2) the analysis module with the Finite Element Method and kinematics analysis, (3) the data access module to store and retrieve the forming information, (4) the communication module, (5) the display module, and (6) the measurement module. The system is useful, irrespective of the heating sources, i.e. flame/gas, laser, or high frequency induction heating. A newly developed automated line heating machine is connected to the proposed system. Experiments and discussions follow.

• Journal of Institute of Control, Robotics and Systems
• /
• v.3 no.2
• /
• pp.179-186
• /
• 1997

In this paper we suggest a development of surface defect inspection algorithms for cold mill strip. The defects which exist in a surface of cold mill strip have a scattering or singular distribution. This paper consists of preprocessing, feature extraction and defect classification. By preprocessing, the binarized defect image is achieved. In this procedure, Top-hit transform, adaptive thresholding, thinning and noise rejection are used. Especially, Top-hit transform using local min/max operation diminishes the effect of bad lighting. In feature extraction, geometric, moment and co-occurrence matrix features are calculated. For the defect classification, multilayer neural network is used. The proposed algorithm showed 15% error rate.

• Journal of Industrial Convergence
• /
• v.18 no.3
• /
• pp.53-61
• /
• 2020

Travel speed is an important parameter for measuring road traffic and incident detection system. In this paper I suggests a model developed for estimating reliable and accurate average roadway link travel speeds using image processing sensor. This method extracts the vehicles from the video image from CCTV, tracks the moving vehicles using deep neural network, and extracts traffic information such as link travel speeds and volume. The algorithm estimates link travel speeds using a robust data-fusion procedure to provide accurate link travel speeds and traffic information to the public. In the field tests, the new model performed better than existing methods.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.25 no.1
• /
• pp.97-104
• /
• 2015

An 'associative cube', a class of auto-associative memories, is revisited here, in which training data and hidden orthogonal basis functions such as wavelet packets or Fourier kernels, are combined in the weight cube. This weight cube has hidden units in its depth, represented by a three dimensional cubic structure. We develop an unsupervised incremental learning mechanism based upon the adaptive least squares method. Training data are mapped into orthogonal basis vectors in a least-squares sense by updating the weights which minimize an energy function. Therefore, a prescribed orthogonal kernel is incrementally assigned to an incoming data. Next, we show how a decoding procedure finds the closest one with a competitive network in the hidden layer. As noisy test data are applied to an associative cube, the nearest one among the original training data are restored in an optimal sense. The simulation results confirm robustness of associative cubes even if test data are heavily distorted by various types of noise.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1995.10a
• /
• pp.393-396
• /
• 1995

In this syudy, the researches classifying the artificial and natural flaws in welding parts are performed using the smart pattern recognition technology. For this purpose the smart signal pattern recognition package including the user defined function was developed and the total procedure including the digital signal processing,feature extraction , feature selection and classifier selection is treated by bulk. Specially it is composed with and discussed using the statistical classifier such as the linear disciminant function classifier, the empirical Bayesian classifier. Also, the smart pattern recognition technology is applied to classification problem of natural flaw(i.e multiple classification problem-crack,lack of penetration,lack of fusion,porosity,and slag inclusion, the planar and volumetric flaw classification problem). According to this results, if appropriately learned the neural network classifier is better than ststistical classifier in the classification problem of natural flaw. And it is possible to acquire the recognition rate of 80% above through it is different a little according to domain extracting the feature and the classifier.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.24 no.10 s.181
• /
• pp.2589-2596
• /
• 2000

As current manufacturing processes require high spindle speed and precise machining, increasing accuracy by reducing volumetric errors of the machine itself, particularly thermal errors, is very important. Thermal errors can be estimated by many empirical models, for example, an FEM model, a neural network model, a linear regression model, an engineering judgment model, etc. This paper discusses to make a modeling of thermal errors efficiently through backward elimination and fuzzy logic strategy. The model of a thermal error using fuzzy logic strategy overcomes limitation of accuracy in the linear regression model or the engineering judgment model. It shows that the fuzzy model has more better performance than linear regression model, though it has less number of thermal variables than the other. The fuzzy model does not need to have complex procedure such like multi-regression and to know the characteristics of the plant, and the parameters of the model can be mathematically calculated. Also, the fuzzy model can be applied to any machine, but it delivers greater accuracy and robustness.

• Journal of Institute of Control, Robotics and Systems
• /
• v.4 no.4
• /
• pp.487-493
• /
• 1998

Camera calibration is the process of determining the coordinate relationship between a camera image and its real world space. Accurate calibration of a camera is necessary for the applications that involve quantitative measurement of camera images. However, if the camera plane is parallel or near parallel to the calibration board on which 2 dimensional objects are defined(this is called "ill-conditioned"), existing solution procedures are not well applied. In this paper, we propose a neural network-based approach to camera calibration for 2D images formed by a mono-camera or a pair of cameras. Multi-layer perceptrons are developed to transform the coordinates of each image point to the world coordinates. The validity of the approach is tested with data points which cover the whole 2D space concerned. Experimental results for both mono-camera and stereo-camera cases indicate that the proposed approach is comparable to Tsai's method［8］. Especially for the stereo camera case, the approach works better than the Tsai's method as the angle between the camera optical axis and the Z-axis increases. Therefore, we believe the approach could be an alternative solution procedure for the ill -conditioned camera calibration.libration.

• Proceedings of the Korean Information Science Society Conference
• /
• 1998.10c
• /
• pp.384-386
• /
• 1998

The use of hand gesture provides an attractive alternative to cumbersome interface devices for human-computer interaction(HCI). Many methods hand gesture recognition using visual analysis have been proposed such as syntactical analysis, neural network(NN), Hidden Markov Model(HMM) and so on. In our research, a HMMs is proposed for alphabetical hand gesture recognition. In the preprocessing stage, the proposed approach consists of three different procedures for hand localization, hand tracking and gesture spotting. The hand location procedure detects the candidated regions on the basis of skin-color and motion in an image by using a color histogram matching and time-varying edge difference techniques. The hand tracking algorithm finds the centroid of a moving hand region, connect those centroids, and thus, produces a trajectory. The spotting a feature database, the proposed approach use the mesh feature code for codebook of HMM. In our experiments, 1300 alphabetical and 1300 untrained gestures are used for training and testing, respectively. Those experimental results demonstrate that the proposed approach yields a higher and satisfying recognition rate for the images with different sizes, shapes and skew angles.

• The KSFM Journal of Fluid Machinery
• /
• v.13 no.6
• /
• pp.42-50
• /
• 2010

The present analysis deals with a numerical procedure for optimizing the shape of elliptical dimples in a cooling channel. The three-dimensional Reynolds-averaged Navier-Stokes (RANS) analysis is employed in conjunction with the SST model for predictions of the turbulent flow and the heat transfer. Three non-dimensional geometric design variables, such as the ellipse dimple diameter ratio, ratio of the dimple depth to the average diameter, and ratio of the distance between dimples to the pitch are considered in the optimization. Twenty-one experimental points within design space are selected by Latin Hypercube Sampling. Each objective function values at these points are evaluated by RANS analysis and producing optimal point using surrogate model. The linear combination of heat transfer coefficient and friction loss related terms with a weighting factor is defined as the objective function. The results show that the optimized elliptical dimple shape improves considerably the heat transfer performance than the circular dimple shape.