• 자원환경지질
• /
• 제41권2호
• /
• pp.219-223
• /
• 2008

해양에서 관측되는 해양지구물리 탐사자료에는 위치오차, 기기오차, 관측오차, 해상 상태 등 다양한 원인에 기인하는 오차가 포함되어 있다. 이에 의해 한 기관에서 해양지구물리 탐사 자료를 취득할 때나 여러 기관에서 취득된 해양지구물리 탐사자료를 취합할 때 많은 교차점오차가 발생하고, 이러한 교차점오차는 부적절한 해석을 야기하는 인위적인 이 상대를 만든다. 교차점오차를 줄이기 위한 다양한 방법들이 제시되었지만, 이들 대부분의 방법들은 교차점을 찾기 위해 각각의 점자료(point data) 혹은 선분자료(segment data)를 모두 비교함으로써, 불필요하게 많은 계산시간을 요구하게 된다. 따라서 본 연구에서는 중복구역나눔 방법을 도입하여 빠르게 교차점을 찾고, 가중치선형내삽 방법을 이용하여 교차점오차를 보정하는 포트란(FORTRAN) 프로그램 (FastXcorr)을 개발하였다.

• 한국정보전자통신기술학회논문지
• /
• 제2권2호
• /
• pp.23-28
• /
• 2009

본 연구에서는 뇌파를 AR모델로 모델링하여 선형예측계수를 특징 파라미터로 이용할 때와 뇌파의 주파수 대역별 상호상관계수를 이용할 때의 감정상태 분류 성능을 비교해 보고자 하였다. 이를 위하여 분노, 슬픔, 기쁨, 안정의 4가지 감정상태에 따른 뇌파를 4개 채널로부터 수집하여 선형예측계수와 ${\theta}$, ${\alpha}$, ${\beta}$ 대역의 주파수 영역에서의 상호상관계수를 추출하여 이들을 특징 파라미터로 한 감정상태 분류 실험을 수행함으로써 두 방법의 감정상태 분류 성능을 비교하였고, 패턴 분류기로는 신경회로망을 이용하였다. 감정 분류 실험 결과 뇌파의 특징 파라미터로서 선형예측계수를 이용한 결과가 상호상관계수를 이용할 때보다 성능이 월등히 좋은 것을 알 수 있었다.

• 시스템엔지니어링학술지
• /
• 제16권1호
• /
• pp.58-67
• /
• 2020

A big data processing method to predict solar power generation using systems engineering approach is developed in this work. For developing analytical method, linear model (LM), support vector machine (SVN), and artificial neural network (ANN) technique are chosen. As evaluation indices, the cross-correlation and the mean square root of prediction error (RMSEP) are used. From multi-variable comparison test, it was found that ANN methodology provides the highest correlation and the lowest RMSEP.

• Journal of Electrical Engineering and Technology
• /
• 제1권4호
• /
• pp.482-489
• /
• 2006

This paper presents the generalized dynamic modeling of self-excited induction generator (SEIG) using state-space approach. The proposed dynamic model consists of induction generator; self-excitation capacitance and load model are expressed in stationary d-q reference frame with the actual saturation curve of the machine. An artificial neural network model is implemented to estimate the machine magnetizing inductance based on the knowledge of magnetizing current. The dynamic performance of SEIG is investigated under no load, with the load, perturbation of load, short circuit at stator terminals, and variation of prime mover speed, variation of capacitance value by considering the effect of main and cross-flux saturation. During voltage buildup the variation in magnetizing inductance is taken into consideration. The performance of SEIG system under various conditions as mentioned above is simulated using MATLAB/SIMULINK and the simulation results demonstrates the feasibility of the proposed system.

• 대한전기학회논문지
• /
• 제43권2호
• /
• pp.329-338
• /
• 1994

This paper presents a novel method of path planning for a quadruped walking robot on irregular terrain. In the previous study on the path planning problem of mobile robots, it has been usually focused on the collision-free path planning for wheeled robots. The path planning problem of legged roboth, however, has unique aspects from the point of viw that the legged robot can cross over the obstacles and the gait constraint should be considered in the process of planning a path. To resolve this unique problem systematically, a new concept of the artificial intensity field of light is numerically constructed over the configuration space of the robot including the transformed obstacles and a feasible path is sought in the field. Also, the efficiency of the proposed method is shown by various simulation results.

• 대한전기학회논문지:전력기술부문A
• /
• 제48권8호
• /
• pp.1014-1021
• /
• 1999

PID controllers with constant gains have been widely used in various control systems. But it is difficult to have uniformly good control performance in all operating conditions. In this paper, we propose a variable PID controller for robot manipulators. We divide total workspace of manipulators into several subspaces. PID controllers in each subspace are optimized using evolution strategy which is a kind of global search algorithm. In real operation, the desired trajectories may cross several subspaces and we select the corresponding gains in each subspace. The gains may have large difference on the boundary of subspaces, which may cause oscillatory motion. So we use artificial neural network to have continuous smooth gain curves to reduce the oscillatory motion. From the experimental results, although the proposed variable PID controller for robots should pay for some computational burden, we have found that the controller is more superior to the conventional constant gain PID controller.

• Geomechanics and Engineering
• /
• 제1권4호
• /
• pp.307-321
• /
• 2009

The paper employs a feed forward neural network with back-propagation algorithm for modeling time dependent swell in clays containing carbonate in the presence of sulfuric acid. The oedometer swell percent is estimated at a nominal surcharge pressure of 6.25 kPa to develop 612 data sets for modeling. The input parameters used in the network include time, sulfuric acid concentration, carbonate percentage, and liquid limit. Among the total data sets, 280 (46%) were assigned to training, 175 (29%) for testing and the remaining 157 data sets (25%) were relegated to cross validation. The network was programmed to process this information and predict the percent swell at any time, knowing the variable involved. The study demonstrates that it is possible to develop a general BPNN model that can predict time dependent swell with relatively high accuracy with observed data ($R^2$=0.9986). The obtained results are also compared with generated non-linear regression model.

• Journal of Smart Tourism
• /
• 제1권4호
• /
• pp.9-18
• /
• 2021

The impacts that AI and robotics systems can and will have on our everyday lives are already making themselves manifest. However, there is a lack of research on the ethical impacts and means for amelioration regarding AI and robotics within tourism and hospitality. Given the importance of designing technologies that cross national boundaries, and given that the tourism and hospitality industry is fundamentally predicated on multicultural interactions, this is an area of research and application that requires particular attention. Specifically, tourism and hospitality have a range of context-unique stakeholders that need to be accounted for in the salient design of AI systems is to be achieved. This paper adopts a stakeholder approach to develop the conceptual framework to centralize human values in designing and deploying AI and robotics systems in tourism and hospitality. The conceptual framework includes several layers - 'Human-human-AI' interaction level, direct and indirect stakeholders, and the macroenvironment. The ethical issues on each layer are outlined as well as some possible solutions to them. Additionally, the paper develops a research agenda on the topic.

• 한국인공지능학회지
• /
• 제5권1호
• /
• pp.29-37
• /
• 2017

The range of problems that can be handled by the activation of big data and the development of hardware has been rapidly expanded and machine learning such as deep learning has become a very versatile technology. In this paper, mnist data set is used as experimental data, and the Cross Entropy function is used as a loss model for evaluating the efficiency of machine learning, and the value of the loss function in the steepest descent method is We applied the Gradient Descent Optimize algorithm to minimize and updated weight and bias via backpropagation. In this way we analyze optimal reliability value corresponding to the number of exercises and optimal reliability value without overfitting. And comparing the overfitting time according to the number of data changes based on the number of training times, when the training frequency was 1110 times, we obtained the result of 92%, which is the optimal reliability value without overfitting.

• Steel and Composite Structures
• /
• 제24권5호
• /
• pp.635-641
• /
• 2017

Safety service improvement and development of efficient maintenance strategies for corroded steel structures are undeniably essential. Therefore, understanding the influence of damage caused by corrosion on the remaining load-carrying capacities such as tensile strength is required. In this study, artificial neural network (ANN) approach is proposed in order to produce a simple, accurate, and inexpensive method developed by using tensile test results, material properties and finite element method (FEM) results to train the ANN model. Initially in reproducing corroded model process, FEM was used to obtain tensile strength of artificial corroded plates, for which surface is developed by a spatial autocorrelation model. By using the corroded surface data and material properties as input data, with tensile strength as the output data, the ANN model could be trained. The accuracy of the ANN result was then verified by using leave-one-out cross-validation (LOOCV). As a result, it was confirmed that the accuracy of the ANN approach and the final output equation was developed for predicting tensile strength without tensile test results and FEM in further work. Though previous studies have been conducted, the accuracy results are still lower than the proposed ANN approach. Hence, the proposed ANN model now enables us to have a simple, rapid, and inexpensive method to predict residual tensile strength more accurately due to corrosion in steel structures.