• Proceedings of the KIEE Conference
• /
• 2006.07d
• /
• pp.1921-1922
• /
• 2006

청소 로봇의 장애물 판단은 청소 로봇이 정확하고 빠르게 장애물을 파악하여 정밀한 제어를 수행하며 청소 효율을 향상 시키는데 중요하다. 청소 로봇이 장애물을 판단하는데 여러 가지 알고리즘이 있지만 신경망 알고리즘 특히, BP(Back-Propagation) 알고리즘을 적용하여 장애물 인식에 있어 반복학습 시키면 청소 로봇은 보다 빠르고 정착하게 장애물을 스스로 판단 할 수 있다. 본 논문에서는 청소 로봇에 부착된 초음파 센서와 장애물과의 거리데이터를 얻어, 이를 BP 알고리즘에 적용하는 것을 연구하며 학습률, 반복학습, 최대 제곱 오차값를 조정한 실험결과로 특성변화를 관찰하고 해석하여 검증한다.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.10
• /
• pp.63-71
• /
• 1996

This paper proposes a method which analyzes facial shape and extracts positions of eyes regardless of the tilt and the size of input iamge. With the extracted feature parameters of facial element by the method, full human faces are recognized by a neural network which BP algorithm is applied on. Input image is changed into binary codes, and then labelled. Area, circumference, and circular degree of the labelled binary image are obtained by using chain code and defined as feature parameters of face image. We first extract two eyes from the similarity and distance of feature parameter of each facial element, and then input face image is corrected by standardizing on two extracted eyes. After a mask is genrated line historgram is applied to finding the feature points of facial elements. Distances and angles between the feature points are used as parameters to recognize full face. To show the validity learning algorithm. We confirmed that the proposed algorithm shows 100% recognition rate on both learned and non-learned data for 20 persons.

• KIEE International Transactions on Electrophysics and Applications
• /
• v.5C no.6
• /
• pp.223-227
• /
• 2005

Insulation failure of traction motor stator coil depends on the continuous stress imposed on it and knowing its insulation condition is an issue of significance for proper safety operation. In this paper, application of the NN (Neural Network) as a scheme of the off-line PD (partial discharge) diagnosis method that occurs at the stator coil of a traction motor was studied. For PD data acquisition, three defective models were made; internal void discharge model, slot discharge model and surface discharge model. PD data for recognition were acquired from a PD detector. Statistical distributions and parameters were calculated to perform recognition between model discharge sources. These statistical distribution parameters are applied to classify PD sources by the NN with a good recognition rate on the discharge sources.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2005.07a
• /
• pp.566-567
• /
• 2005

Until now variable pattern classification methods have been introduced. So, variable methods in PD source classification were applied. NN(neural network) the most used scheme as a PD(partial discharge) source classification. But in recent year another method were developed. These methods is present superior to NN in the field of image and signal process function of classification. In this paper, it is show classification result in PD source using three methods; that is, BP(back-propagation), ANFIS(adaptive neuro-fuzzy inference system), PCA-LDA(principle component analysis-linear discriminant analysis).

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2005.07a
• /
• pp.221-222
• /
• 2005

In this paper, application of NN (Neural Network) as a method of pattern discrimination of PD(partial discharge) which occurs at the stator coil of traction motor was studied. For PD data acquisition, three defective models are manufactured such as internal discharge model, slot discharge model and surface discharge model. PD data for recognition were acquired from PD detector and DAQ board which is able to analysis the PD signal and perform the pattern discrimination. Statistical distributions and parameters are calculated to discriminate PD sources. And also these statistical distribution parameters are applied to classify PD sources by BP and has good recognition rate on the discharge sources.

• Journal of Information Processing Systems
• /
• v.20 no.3
• /
• pp.409-417
• /
• 2024

Excellent apparel design can increase market competitiveness. This article briefly introduced the theory of fractals and its application in the field of apparel design. The convolutional neural network (CNN) algorithm was used to assist in the evaluation of apparel designs. In the case analysis, the accuracy of the evaluation was validated by comparing the CNN algorithm with two other intelligent algorithms, support vector machine (SVM) and back propagation (BP). The evaluation of the proposed design showed that compared with SVM and BP algorithms, the CNN algorithm had higher accuracy in evaluating apparel designs. The evaluation result of the proposed apparel design not only further verifies the effectiveness of the CNN algorithm, but also demonstrates that the theory of fractals can be effectively applied in apparel design to provide more innovative designs.

• Journal of Biosystems Engineering
• /
• v.24 no.2
• /
• pp.153-158
• /
• 1999

As a basic research for the development of the automatic grading and sorting system for dried oak mushrooms, the device to acquire both cap and gill side images of mushroom has been developed and neural network based side recognition and quality grading has been proposed via inputting both side images. 20 quality grades have been selected considering the requirement of grade classifications imposed by the mushroom company. Developed DC motor driven‘V’type reversing device for the image acquisition of both side images of mushroom showed more than 95% success. Most error was caused by very small size mushrooms with a radius of around 1cm. However, it required a further research to reduce the reversing time. Grading and side recognition were performed via inputting normalized size factors and average gray levels of $8{\times}8$ grids converted from the raw images of both surfaces to the multi-layer back propagation(BP) network. Accuracy of the grading showed about 88.5% and the total grading time including reversing operation was around 2 seconds.

• Journal of the Korean Society for Precision Engineering
• /
• v.9 no.4
• /
• pp.22-35
• /
• 1992

In highly automated machining centers, intelligent sensor fddeback systems are indispensable on order to monitor their operations, to ensure efficient metal removal, and to initate remedial action in the event of accident. In this study, an on-line tool wear detection system for thrning operations is developed, and experimentally evaluated. The system employs multiple sensors and the signals from these sensors are processed using a multichannel autoegressive (AR) series model. The resulting output from the signal processing block is then fed to a previously tranied artificial neural network (multiayered perceptron) to make a final decision on the state of the cutting tool. To learn the necessary input/output mapping for tool wear detection, the weithts and thresholds of the network are adjusted according to the back propagation (BP) method during off-line training. The results of experimental evaluation show that the system works well over a wide range of cutting conditions, and the ability of the system to detect tool wear is improved due to the generalization, fault-tolearant and self-ofganizing properties of the neural network.

• Journal of Korea Society of Industrial Information Systems
• /
• v.11 no.1
• /
• pp.77-85
• /
• 2006

In this paper, we propose a hierarchical classifier of object images using neural networks for content-based image classification. The images for classification are object images that can be divided into foreground and background. In the preprocessing step, we extract the object region and shape-based texture features extracted from wavelet transformed images. We group the image classes into clusters which have similar texture features using Principal Component Analysis(PCA) and K-means. The hierarchical classifier has five layes which combine the clusters. The hierarchical classifier consists of 59 neural network classifiers learned with the back propagation algorithm. Among the various texture features, the diagonal moment was the most effective. A test with 1000 training data and 1000 test data composed of 10 images from each of 100 classes shows classification rates of 81.5% and 75.1% correct, respectively.

• Journal of Advanced Marine Engineering and Technology
• /
• v.27 no.1
• /
• pp.24-32
• /
• 2003

The PID controller has been widely used in industrial applications due to its simple structure and robustness. Even if it is initially well tuned, the PID controller must be retuned to maintain acceptable performance when there are system parameter changes due to the change of operation conditions. In this paper, a self-tuning control scheme which comprises a parameter estimator, a NN-based rule emulator and a PID controller is proposed, which can cope with changing environments. This method involves combining neural networks and real-coded genetic algorithms(RCGAs) with conventional approaches to provide a stable and satisfactory response. A RCGA-based parameter estimation method is first described to obtain the first-order with time delay model from over-damped high-order systems. Then, a set of optimum PID parameters are calculated based on the estimated model such that they cover the entire spectrum of system operations and an optimum tuning rule is trained with a BP-based neural network. A set of simulation works on systems with time delay are carried out to demonstrate the effectiveness of the proposed method.