• Journal of Korea Society of Industrial Information Systems
• /
• v.22 no.5
• /
• pp.17-24
• /
• 2017

Recently Deep-learning, Technologies using Large or Deep Artificial Neural Networks, have Shown Remarkable Performance, and the Increasing Size of the Network Contributes to its Performance Improvement. However, the Increase in the Size of the Neural Network Leads to an Increase in the Calculation Amount, which Causes Problems Such as Circuit Complexity, Price, Heat Generation, and Real-time Restriction. In This Paper, We Propose and Test a Method to Reduce the Number of Network Connections by Effectively Pruning the Redundancy in the Connection and Showing the Difference between the Performance and the Desired Range of the Original Neural Network. In Particular, we Proposed a Simple Method to Improve the Performance by Re-learning and to Guarantee the Desired Performance by Allocating the Error Rate per Layer in Order to Consider the Difference of each Layer. Experiments have been Performed on a Typical Neural Network Structure such as FCN (full connection network) and CNN (convolution neural network) Structure and Confirmed that the Performance Similar to that of the Original Neural Network can be Obtained by Only about 1/10 Connection.

• Journal of the Korea Society of Computer and Information
• /
• v.24 no.5
• /
• pp.149-156
• /
• 2019

The existing subtract image based intrusion detection system for CCTV digital images has a problem that it can not distinguish intruders from moving backgrounds that exist in the natural environment. In this paper, we tried to solve the problems of existing system by designing real - time intrusion detection system for CCTV digital image by combining subtract image based intrusion detection method and background learning artificial neural network technology. Our proposed system consists of three steps: subtract image based intrusion detection, background artificial neural network learning stage, and background artificial neural network evaluation stage. The final intrusion detection result is a combination of result of the subtract image based intrusion detection and the final intrusion detection result of the background artificial neural network. The step of subtract image based intrusion detection is a step of determining the occurrence of intrusion by obtaining a difference image between the background cumulative average image and the current frame image. In the background artificial neural network learning, the background is learned in a situation in which no intrusion occurs, and it is learned by dividing into a detection window unit set by the user. In the background artificial neural network evaluation, the learned background artificial neural network is used to produce background recognition or intrusion detection in the detection window unit. The proposed background learning intrusion detection system is able to detect intrusion more precisely than existing subtract image based intrusion detection system and adaptively execute machine learning on the background so that it can be operated as highly practical intrusion detection system.

• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.77-91
• /
• 2022

Various monitoring systems have been implemented in civil infrastructure to ensure structural safety and integrity. In long-term monitoring, these systems generate a large amount of data, where anomalies are not unusual and can pose unique challenges for structural health monitoring applications, such as system identification and damage detection. Therefore, developing efficient techniques is quite essential to recognize the anomalies in monitoring data. In this study, several machine learning techniques are explored and implemented to detect and classify various types of data anomalies. A field dataset, which consists of one month long acceleration data obtained from a long-span cable-stayed bridge in China, is employed to examine the machine learning techniques for automated data anomaly detection. These techniques include the statistic-based pattern recognition network, spectrogram-based convolutional neural network, image-based time history convolutional neural network, image-based time-frequency hybrid convolution neural network (GoogLeNet), and proposed ensemble neural network model. The ensemble model deliberately combines different machine learning models to enhance anomaly classification performance. The results show that all these techniques can successfully detect and classify six types of data anomalies (i.e., missing, minor, outlier, square, trend, drift). Moreover, both image-based time history convolutional neural network and GoogLeNet are further investigated for the capability of autonomous online anomaly classification and found to effectively classify anomalies with decent performance. As seen in comparison with accuracy, the proposed ensemble neural network model outperforms the other three machine learning techniques. This study also evaluates the proposed ensemble neural network model to a blind test dataset. As found in the results, this ensemble model is effective for data anomaly detection and applicable for the signal characteristics changing over time.

• International Journal of Control, Automation, and Systems
• /
• v.3 no.2
• /
• pp.202-210
• /
• 2005

This paper presents a new neural network control for nonlinear bridge systems with earthquake excitation. We design multi-layer neural network controllers with a single hidden layer. The selection of an optimal number of neurons in the hidden layer is an important design step for control performance. To select an optimal number of hidden neurons, we progressively add one hidden neuron and observe the change in a performance measure given by the weighted sum of the system error and the control force. The number of hidden neurons which minimizes the performance measure is selected for implementation. A neural network was trained for mitigating vibrations of bridge systems caused by El Centro earthquake. We applied the proposed control approach to a single-degree-of-freedom (SDOF) and a two-degree-of-freedom (TDOF) bridge system. We assessed the robustness of the control system using randomly generated earthquake excitations which were not used in training the neural network. Our results show that the neural network controller drastically mitigates the effect of the disturbance.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.1478-1481
• /
• 2005

As the model of most practical system cannot be obtained, the practice of typical control method is limited. Accordingly, numerous artificial intelligence control methods have been used widely. Fuzzy control and neural network control have been an important point in the developing process of the field. This paper is proposed adaptive fuzzy-neural network based on the vector controlled interior permanent magnet synchronous motor drive system. The fuzzy-neural network is first utilized for the speed control. A model reference adaptive scheme is then proposed in which the adaptation mechanism is executed using fuzzy-neural network. Also, this paper is proposed estimation of speed of interior permanent magnet synchronous motor using artificial neural network controller. The back-propagation neural network technique is used to provide a real time adaptive estimation of the motor speed. The error between the desired state variable and the actual one is back-propagated to adjust the rotor speed, so that the actual state variable will coincide with the desired one. The back-propagation mechanism is easy to derive and the estimated speed tracks precisely the actual motor speed. This paper is proposed the analysis results to verify the effectiveness of the new method.

• Journal of Institute of Control, Robotics and Systems
• /
• v.16 no.10
• /
• pp.963-968
• /
• 2010

This paper handles ZMP based control that is inspired by neural networks for humanoid robot walking on varying sloped surfaces. Humanoid robots are currently one of the most exciting research topics in the field of robotics, and maintaining stability while they are standing, walking or moving is a key concern. To ensure a steady and smooth walking gait of such robots, a feedforward type of neural network architecture, trained by the back propagation algorithm is employed. The inputs and outputs of the neural network architecture are the ZMPx and ZMPy errors of the robot, and the x, y positions of the robot, respectively. The neural network developed allows the controller to generate the desired balance of the robot positions, resulting in a steady gait for the robot as it moves around on a flat floor, and when it is descending slope. In this paper, experiments of humanoid robot walking are carried out, in which the actual position data from a prototype robot are measured in real time situations, and fed into a neural network inspired controller designed for stable bipedal walking.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.58 no.1
• /
• pp.126-131
• /
• 2009

This study was carried out for the reliability of PD(Partial Discharge) pattern recognition. For the pattern recognition, the database for PD was established by use of self-designed insulation defects which occur and were mostly critical in GIS(Gas Insulated Switchgear). The acquired database was analyzed to distinguish patterns by means of PRPD(Phase Resolved Partial Discharge) method and stored to the form with to unite the average amplitude of PD pulse and the number of PD pulse as the input data of neural network. In order to prove the performance of genetic algorithm combined with neural network, the neural networks with trial-and-error method and the neural network with genetic algorithm were trained by same training data and compared to the results of their pattern recognition rate. As a result, the recognition success rate of defects was 93.2% and the neural network train process by use of trial-and-error method was very time consuming. The recognition success rate of defects, on the other hand, was 100% by applying the genetic algorithm at neural network and it took a relatively short time to find the best solution of parameters for optimization. Especially, it could be possible that the scrupulous parameters were obtained by genetic algorithm.

• Journal of Digital Convergence
• /
• v.15 no.3
• /
• pp.201-209
• /
• 2017

This paper is a smartphone object recognition system using hierarchical convolutional neural network. The overall configuration is a method of communicating object information to the smartphone by matching the collected data by connecting the smartphone and the server and recognizing the object to the convergence neural network in the server. It is also compared to a hierarchical convolutional neural network and a fractional convolutional neural network. Hierarchical convolutional neural networks have 88% accuracy, fractional convolutional neural networks have 73% accuracy and 15%p performance improvement. Based on this, it shows possibility of expansion of T-Commerce market connected with smartphone and broadcasting media.

• Journal of Electrical Engineering and Technology
• /
• v.13 no.4
• /
• pp.1724-1731
• /
• 2018

The conventional polynomial neural network (PNN) is a classical flexible neural structure and self-organizing network, however it is not free from the limitation of overfitting problem. In this study, we propose a space search-optimized polynomial neural network (ssPNN) structure to alleviate this problem. Ranking selection is realized by means of ranking selection-based performance index (RS_PI) which is combined with conventional performance index (PI) and coefficients based performance index (CPI) (viz. the sum of squared coefficient). Unlike the conventional PNN, L2-norm regularization method for estimating the polynomial coefficients is also used when designing the ssPNN. Furthermore, space search optimization (SSO) is exploited here to optimize the parameters of ssPNN (viz. the number of input variables, which variables will be selected as input variables, and the type of polynomial). Experimental results show that the proposed ranking selection-based polynomial neural network gives rise to better performance in comparison with the neuron fuzzy models reported in the literatures.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.49 no.9
• /
• pp.512-519
• /
• 2000

A Neural networks has been used for a expert system and fault diagnosis system. It is possible to nonlinear function mapping and parallel processing. Therefore It has been developing for a Diagnosis system of nuclear plower plant. In general Neural Networks is a static mapping but Dynamic Neural Network(DNN) is dynamic mapping.쪼두 a fault occur in system a state of system is changed with transient state. Because of a previous state signal is considered as a information DNN is better suited for diagnosis systems than static neural network. But a DNN has many weights so a real time implementation of diagnosis system is in need of a rapid network architecture. This paper presents a algorithm for RCP monitoring Alarm diagnosis system using Self Dynamic Neural Network(SDNN). SDNN has considerably fewer weights than a general DNN. Since there is no interlink among the hidden layer. The effectiveness of Alarm diagnosis system using the proposed algorithm is demonstrated by applying to RCP monitoring in Nuclear power plant.