• Journal of the Korean Geotechnical Society
• /
• v.16 no.1
• /
• pp.75-81
• /
• 2000

The anisotropy of soils has an important effect on stress-strain behavior. In this study, an attempt has been made to implement artificial neural network model for modeling the stress-strain relationship and predicting the undrained shear strength of normally consolidated clay with varying consolidation pressure ratios. The multi-layer neural network model, adopted in this study, utilizes the error back-propagation loaming algorithm. The artificial neural networks use the results of undrained triaxial test with various consolidation pressure ratios and different effective vertical consolidation pressure fur learning and testing data. After learning from a set of actual laboratory testing data, the neural network model predictions of the undrained shear strength of the normally consolidated clay are found to agree well with actual measurements. The predicted values by the artificial neural network model have a determination coefficient$(r^2)$ above 0.973 compared with the measured data. Therefore, this results show a positive potential for the applications of well-trained neural network model in predicting the undrained shear strength of cohesive soils.

• Journal of Sensor Science and Technology
• /
• v.12 no.6
• /
• pp.289-297
• /
• 2003

In this paper, we propose accurate classification method of an EEG signals during a mental tasks. In the experimental task, subjects achieved through the process of responding to visual stimulus, understanding the given problem, controlling hand motions, and select a key. To recognize the subjects' selection time, we analyzed with 4 types feature from the filtered brain waves at frequency bands of $\alpha$, $\beta$, $\theta$, $\gamma$ waves. From the analysed features, we construct specific rules for each subject meta rules including common factors in all subjects. In this system, the architecture of the neural network is a three layered feedforward networks with one hidden layer which implements the error back propagation learning algorithm. Applying the algorithms to 4 subjects show 87% classification success rates. In this paper, the proposed detection method can be a basic technology for brain-computer-interface by combining with discrimination methods.

• The KIPS Transactions:PartB
• /
• v.13B no.6 s.109
• /
• pp.615-624
• /
• 2006

The objective of this paper is to implement parallel multi-layer ANN(Artificial Neural Network) simulator based on the mobile agent system which is executed in parallel in the virtual parallel distributed computing environment. The Multi-Layer Neural Network is classified by training session, training data layer, node, md weight in the parallelization-level. In this study, We have developed and evaluated the simulator with which it is feasible to parallel the ANN in the training session and training data parallelization because these have relatively few network traffic. In this results, we have verified that the performance of parallelization is high about 3.3 times in the training session and training data. The great significance of this paper is that the performance of ANN's execution on virtual parallel computer is similar to that of ANN's execution on existing super-computer. Therefore, we think that the virtual parallel computer can be considerably helpful in developing the neural network because it decreases the training time which needs extra-time.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.480-480
• /
• 2000

In this paper, we proposed a fuzzy-neuro controller to control the speed of wound rotor induction motor with slip energy recovery. The speed is limited at some range of sub-synchronous speed of the rotating magnetic field. Control speed by adjusting resistance value in the rotor circuit that occurs the efficiency of power are reduced, because of the slip energy is lost when it passes through the rotor resistance. The control system is designed to maintain efficiency of motor. Recently, the emergence of artificial neural networks has made it conductive to integrate fuzzy controllers and neural models for the development of fuzzy control systems, Fuzzy-neuro controller has been designed by integrating two neural network models with a basic fuzzy logic controller. Using the back propagation algorithm, the first neural network is trained as a plant emulator and the second neural network is used as a compensator for the basic fuzzy controller to improve its performance on-line. The function of the neural network plant emulator is to provide the correct error signal at the output of the neural fuzzy compensator without the need for any mathematical modeling of the plant. The difficulty of fine-tuning the scale factors and formulating the correct control rules in a basic fuzzy controller may be reduced using the proposed scheme. The scheme is applied to the control speed of a wound rotor induction motor process. The control system is designed to maintain efficiency of motor and compensate power factor of system. That is: the proposed controller gives the controlled system by keeping the speed constant and the good transient response without overshoot can be obtained.

• Journal of Sensor Science and Technology
• /
• v.8 no.6
• /
• pp.461-468
• /
• 1999

In this paper, we have implemented a gas recognition system for classification and identification of explosive gases such as methane, propane, and butane using a sensor array and an artificial neural network. Such explosive gases which can be usually detected in the oil factory and LPG pipeline are very dangerous for a human being. We analyzed the characteristics of a multi-dimensional sensor signals obtained from the nine sensors using the principal component analysis(PCA) technique. Also, we implemented a gas pattern recognizer using a multi-layer neural network with error back propagation learning algorithm, which can classify and identify the sorts of gases and concentrations for each gas. The simulation and experimental results show that the proposed gas recognition system is effective to identify the explosive gases. And also, we used DSP board(TMS320C31) to implement the proposed gas recognition system using the neural network for real time processing.

• The Journal of Korea Robotics Society
• /
• v.7 no.3
• /
• pp.171-181
• /
• 2012

Outdoor mobile robots are faced with various terrain types having different characteristics. To run safely and carry out the mission, mobile robot should recognize terrain types, physical and geometric characteristics and so on. It is essential to control appropriate motion for each terrain characteristics. One way to determine the terrain types is to use non-contact sensor data such as vision and laser sensor. Another way is to use contact sensor data such as slope of body, vibration and current of motor that are reaction data from the ground to the tire. In this paper, we presented experimental results on terrain classification using contact sensor data. We made a mobile robot for collecting contact sensor data and collected data from four terrains we chose for experimental terrains. Through analysis of the collecting data, we suggested a new method of terrain feature extraction considering physical characteristics and confirmed that the proposed method can classify the four terrains that we chose for experimental terrains. We can also be confirmed that terrain feature extraction method using Fast Fourier Transform (FFT) typically used in previous studies and the proposed method have similar classification performance through back propagation learning algorithm. However, both methods differ in the amount of data including terrain feature information. So we defined an index determined by the amount of terrain feature information and classification error rate. And the index can evaluate classification efficiency. We compared the results of each method through the index. The comparison showed that our method is more efficient than the existing method.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.46 no.2
• /
• pp.65-74
• /
• 2009

In optimizing the neural network structure, there are two methods of the pruning scheme and the constructive scheme. In this paper we use the pruning scheme to optimize neural network structure, and the genetic algorithm to find out its optimum node pruning. In the conventional researches, the input and hidden layers were optimized separately. On the contrary we attempted to optimize the two layers simultaneously by encoding two layers in a chromosome. The offspring networks inherit the weights from the parent. For teaming, we used the existing error back-propagation algorithm. In our experiment with various databases from UCI Machine Learning Repository, we could get the optimal performance when the network size was reduced by about $8{\sim}25%$. As a result of t-test the proposed method was shown better performance, compared with other pruning and construction methods through the cross-validation.

• Journal of the Korea Computer Graphics Society
• /
• v.26 no.3
• /
• pp.87-97
• /
• 2020

In this paper, we propose a novel neural relighting that infers a relighted rendering image based on the user-guided specular highlight map. The proposed network utilizes a pre-trained neural renderer as a backbone network learned from the rendered image of a 3D scene with various lighting conditions. We jointly optimize a 3D light position and its associated relighted image by back-propagation, so that the difference between the base image and the relighted image is similar to the user-guided specular highlight map. The proposed method has the advantage of being able to explicitly infer the 3D lighting position, while providing the artists' preferred 2D screen-space interface. The performance of the proposed network was measured under the conditions that can establish ground truths, and the average error rate of light position estimations is 0.11, with the normalized 3D scene size.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.39C no.9
• /
• pp.820-827
• /
• 2014

One of the basic problems in Wireless Sensor Networks (WSNs) is the localization of the sensor nodes based on the known location of numerous anchor nodes. WSNs generally consist of a large number of sensor nodes and recording the location of each sensor nodes becomes a difficult task. On the other hand, based on the application environment, the nodes may be subject to mobility and their location changes with time. Therefore, a scheme that will autonomously estimate or calculate the position of the sensor nodes is desirable. This paper presents an intelligent localization scheme, which is an artificial neural network (ANN) based localization scheme used to estimate the position of the unknown nodes. In the proposed method, three anchors nodes are used. The mobile or deployed sensor nodes request a beacon from the anchor nodes and utilizes the received signal strength indicator (RSSI) of the beacons received. The RSSI values vary depending on the distance between the mobile and the anchor nodes. The three RSSI values are used as the input to the ANN in order to estimate the location of the sensor nodes. A feed-forward artificial neural network with back propagation method for training has been employed. An average Euclidian distance error of 0.70 m has been achieved using a ANN having 3 inputs, two hidden layers, and two outputs (x and y coordinates of the position).

• Journal of Institute of Control, Robotics and Systems
• /
• v.8 no.2
• /
• pp.126-135
• /
• 2002

The study is concerned with an approach to the design of new architectures of fuzzy neural networks and the discussion of comprehensive design methodology supporting their development. We propose an Adaptive Fuzzy Polynomial Neural Networks(APFNN) based on Fuzzy Neural Networks(FNN) and Self-organizing Networks(SON) for model identification of complex and nonlinear systems. The proposed AFPNN is generated from the mutually combined structure of both FNN and SON. The one and the other are considered as the premise and the consequence part of AFPNN, respectively. As the premise structure of AFPNN, FNN uses both the simplified fuzzy inference and error back-propagation teaming rule. The parameters of FNN are refined(optimized) using genetic algorithms(GAs). As the consequence structure of AFPNN, SON is realized by a polynomial type of mapping(linear, quadratic and modified quadratic) between input and output variables. In this study, we introduce two kinds of AFPNN architectures, namely the basic and the modified one. The basic and the modified architectures depend on the number of input variables and the order of polynomial in each layer of consequence structure. Owing to the specific features of two combined architectures, it is possible to consider the nonlinear characteristics of process system and to obtain the better output performance with superb predictive ability. The availability and feasibility of the AFPNN are discussed and illustrated with the aid of two representative numerical examples. The results show that the proposed AFPNN can produce the model with higher accuracy and predictive ability than any other method presented previously.