• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.640-645
• /
• 2004

This paper proposes a direct adaptive control method using self-recurrent wavelet neural network (SRWNN) for stable path tracking of mobile robots. The architecture of the SRWNN is a modified model of the wavelet neural network (WNN). Unlike the WNN, since a mother wavelet layer of the SRWNN is composed of self-feedback neurons, the SRWNN has the ability to store the past information of the wavelet. For this ability of the SRWNN, the SRWNN is used as a controller with simpler structure than the WNN in our on-line control process. The gradient-descent method with adaptive learning rates (ALR) is applied to train the parameters of the SRWNN. The ALR are derived from discrete Lyapunov stability theorem, which are used to guarantee the stable path tracking of mobile robots. Finally, through computer simulations, we demonstrate the effectiveness and stability of the proposed controller.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.316-321
• /
• 2003

This paper presents a design method of the wavelet neural network based controller using direct adaptive control method to deal with a stable intelligent control of chaotic systems. The various uncertainties, such as mechanical parametric variation, external disturbance, and unstructured uncertainty influence the control performance. However, the conventional control methods such as optimal control, adaptive control and robust control may not be feasible when an explicit, faithful mathematical model cannot be constructed. Therefore, an intelligent control system that is an on-line trained WNN controller based on direct adaptive control method with adaptive learning rates is proposed to control chaotic nonlinear systems whose mathematical models are not available. The adaptive learning rates are derived in the sense of discrete-type Lyapunov stability theorem, so that the convergence of the tracking error can be guaranteed in the closed-loop system. In the whole design process, the strict constrained conditions and prior knowledge of the controlled plant are not necessary due to the powerful learning ability of the proposed intelligent control system. The gradient-descent method is used for training a wavelet neural network controller of chaotic systems. Finally, the effectiveness and feasibility of the proposed control method is demonstrated with application to the chaotic systems.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.34S no.7
• /
• pp.102-111
• /
• 1997

In this paper, a patern recognition method of surface mount device(SMD) IC using wavelet transform and neural network is proposed. We chose the feature parameter according to the characteristics of coefficient matrix which is obtained from four level discrete wavelet transform (DWT). These feature parameters are normalized and then used for the input vector of neural network which is capable of adapting the surroundings such as variation of illumination, arrangement of objects and translation. Experimental results show that when the same form of feature pattern, as is used for learning, is put into neural network and gained 100% rate ofrecognition irrespective of SMD IC kinds, location and variation of illumination. In the case of unused feature pattern for learning, the recognition rate is 85.9% under the similar surroundings, where as an average recognition rate is 96.87% for the case of reregulated value of illumination. Proosed method is relatively simple compared with the traditional space domain method in extracting the feature parameter and is also well suited for recognizing the pattern's class, position and existence. It can also shorten the processing tiem better than method extracting feature parameter with the use of discrete cosine transform(DCT) and adapt the surroundings such as variation of illumination, the arrangement and the translation of SMD IC.

• Structural Monitoring and Maintenance
• /
• v.3 no.4
• /
• pp.377-393
• /
• 2016

The present work develops an expert system for detecting and predicting the crude oil types and properties at normal temperature ${\theta}=25^{\circ}C$, by evaluating the dielectric properties of the fluid transfused inside glass fiber reinforced epoxy (GFRE) composite pipelines, by using electrical capacitance sensor (ECS) technique, then used the data measurements from ECS to predict the types of the other crude oil transfused inside the pipeline, by designing an efficient artificial neural network (ANN) architecture. The variation in the dielectric signatures are employed to design an electrical capacitance sensor (ECS) with high sensitivity to detect such problem. ECS consists of 12 electrodes mounted on the outer surface of the pipe. A finite element (FE) simulation model is developed to measure the capacitance values and node potential distribution of ECS electrodes by ANSYS and MATLAB, which are combined to simulate sensor characteristic. Radial Basis neural network (RBNN), structure is applied, trained and tested to predict the finite element (FE) results of crude oil types transfused inside (GFRE) pipe under room temperature using MATLAB neural network toolbox. The FE results are in excellent agreement with an RBNN results, thus validating the accuracy and reliability of the proposed technique.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.1
• /
• pp.64-68
• /
• 2020

In a conventional speech recognition system using Gaussian Mixture Model-Hidden Markov Model (GMM-HMM), the cepstral feature normalization method based on pole filtering was effective in improving the performance of recognition of short utterances in noisy environments. In this paper, the usefulness of this method for the state-of-the-art speech recognition system using Deep Neural Network (DNN) is examined. Experimental results on AURORA 2 DB show that the cepstral mean and variance normalization based on pole filtering improves the recognition performance of very short utterances compared to that without pole filtering, especially when there is a large mismatch between the training and test conditions.

• The Journal of the Convergence on Culture Technology
• /
• v.3 no.4
• /
• pp.171-175
• /
• 2017

Recently, Training of the convolutional neural network (CNN) entails many iterative computations. Therefore, a method of accelerating the training speed through parallel processing using the hardware specifications of GPGPU is actively researched. In this paper, the operations of the feature extraction unit and the classification unit are divided into blocks and threads of GPGPU and processed in parallel. Convolution and Pooling operations of the feature extraction unit are processed in parallel at once without sequentially processing. As a result, proposed method improved the training time about 314%.

• Smart Structures and Systems
• /
• v.18 no.6
• /
• pp.1233-1250
• /
• 2016

Cracks in plate-like structures are some of the main reasons for destruction of the entire structure. In this study, a novel two-stage methodology is proposed for damage detection of flexural plates using an optimized artificial neural network. In the first stage, location of damages in plates is investigated using curvature-moment and curvature-moment derivative concepts. After detecting the damaged areas, the equations for damage severity detection are solved via Bat Algorithm (BA). In the second stage, in order to efficiently reduce the computational cost of model updating during the optimization process of damage severity detection, multiple damage location assurance criterion index based on the frequency change vector of structures are evaluated using properly trained cascade feed-forward neural network (CFNN) as a surrogate model. In order to achieve the most generalized neural network as a surrogate model, its structure is optimized using binary version of BA. To validate this proposed solution method, two examples are presented. The results indicate that after determining the damage location based on curvature-moment derivative concept, the proposed solution method for damage severity detection leads to significant reduction of computational time compared with direct finite element method. Furthermore, integrating BA with the efficient approximation mechanism of finite element model, maintains the acceptable accuracy of damage severity detection.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.16 no.4
• /
• pp.238-245
• /
• 2016

Inverters are considered the basic building blocks of industrial electrical drive systems that are widely used for various applications; however, the failure of electronic switches mainly affects the constancy of these inverters. For safe and reliable operation of an electrical drive system, faults in power electronic switches must be detected by an efficient system that is capable of identifying the type of faults. In this paper, an open switch fault identification technique for a three-phase inverter is presented. Single, double, and triple switching faults can be diagnosed using this method. The detection mechanism is based on stator current analysis. Discrete wavelet transform (DWT) using Daubechies is performed on the Clarke transformed (-) stator current and features are extracted from the wavelets. An artificial neural network is then used for the detection and identification of faults. To prove the feasibility of this method, a Simulink model of the DWT-based feature extraction scheme using a neural network for the proposed fault detection system in a three-phase inverter with an induction motor is briefly discussed with simulation results. The simulation results show that the designed system can detect faults quite efficiently, with the ability to differentiate between single and multiple switching faults.

• KIISE Transactions on Computing Practices
• /
• v.24 no.1
• /
• pp.53-57
• /
• 2018

A poisoning accident is often caused by a situation in which people eat poisonous mushrooms because they cannot distinguish between edible mushrooms and poisonous mushrooms. In this paper, we propose an automatic mushroom recognition system by using the convolutional neural network. We collected 1478 mushroom images of 38 species using image crawling, and used the dataset for learning the convolutional neural network. A comparison experiment using AlexNet, VGGNet, and GoogLeNet was performed using the collected datasets, and a comparison experiment using a class number expansion and a fine-tuning technique for transfer learning were performed. As a result of our experiment, we achieve 82.63% top-1 accuracy and 96.84% top-5 accuracy on test set of our dataset.

• Computers and Concrete
• /
• v.14 no.6
• /
• pp.727-743
• /
• 2014

The management of existing concrete bridges has become a major social concern in many developed countries due to the large number of bridges exhibiting signs of significant deterioration. This problem has increased the demand for effective maintenance and renewal planning. In order to implement an appropriate management procedure for a structure, a wide array of corrective strategies must be evaluated with respect to not only the condition state of each defect but also safety, economy and sustainability. This paper describes a new performance evaluation system for existing concrete bridges. The system evaluates performance based on load carrying capability and durability from the results of a visual inspection and specification data, and describes the necessity of maintenance. It categorizes all girders and slabs as either unsafe, severe deterioration, moderate deterioration, mild deterioration, or safe. The technique employs an expert system with an appropriate knowledge base in the evaluation. A characteristic feature of the system is the use of neural networks to evaluate the performance and facilitate refinement of the knowledge base. The neural network proposed in the present study has the capability to prevent an inference process and knowledge base from becoming a black box. It is very important that the system is capable of detailing how the performance is calculated since the road network represents a huge investment. The effectiveness of the neural network and machine learning method is verified by comparing diagnostic results by bridge experts.