• Proceedings of the KIPE Conference
• /
• 2006.06a
• /
• pp.345-347
• /
• 2006

In this paper, a novel method for wind speed estimation in wind power generation systems is presented. The proposed algorithm is based on estimating the wind speed using Support-Vector-Machines for regression (SVR). The wind speed is estimated using the generator power-speed characteristics as a set of training vectors. SVR is trained off-line to predict a continuos-valued function between the system's inputs and wind speed value. The predicted off-line function as well as the instantaneous generator power and speed are then used to determine the unknown winds speed on-line. The simulation results show that SVR can define the corresponding wind speed rapidly and accurately to determine the optimum generator speed reference for maximum power point tracking.

• Journal of Power Electronics
• /
• v.8 no.4
• /
• pp.345-353
• /
• 2008

In this paper, a novel loss minimization of an induction generator in wind energy generation systems is presented. The proposed algorithm is based on the flux level reduction, for which the generator d-axis current reference is estimated using support vector regression (SVR). Wind speed is employed as an input of the SVR and the samples of the generator d-axis current reference are used as output to train the SVR algorithm off-line. Data samples for wind speed and d-axis current are collected for the training process, which plots a relation of input and output. The predicted off-line function and the instantaneous wind speed are then used to determine the d-axis current reference. It is shown that the effect of loss minimization is more significant at low wind speed and the loss reduction is about to 40% at 4[m/s] wind speed. The validity of the proposed scheme has been verified by experimental results.

• Journal of the Korean Data and Information Science Society
• /
• v.26 no.6
• /
• pp.1537-1545
• /
• 2015

Support vector quantile regression (SVQR) can be obtained by applying support vector machine with a check function instead of an e-insensitive loss function into the quantile regression, which still requires to solve a quadratic program (QP) problem which is time and memory expensive. In this paper we propose an SVQR whose objective function is composed of an asymmetric quadratic loss function. The proposed method overcomes the weak point of the SVQR with the check function. We use the iterative procedure to solve the objective problem. Furthermore, we introduce the generalized cross validation function to select the hyper-parameters which affect the performance of SVQR. Experimental results are then presented, which illustrate the performance of proposed SVQR.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.2
• /
• pp.461-471
• /
• 2017

In this paper we propose a feature selection method identifying important features in the semivarying coefficient model. One important issue in semivarying coefficient model is how to estimate the parametric and nonparametric components. Another issue is how to identify important features in the varying and the constant effects. We propose a feature selection method able to address this issue using generalized cross validation functions of the varying coefficient least squares support vector regression (LS-SVR) and the linear LS-SVR. Numerical studies indicate that the proposed method is quite effective in identifying important features in the varying and the constant effects in the semivarying coefficient model.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.15 no.4
• /
• pp.309-318
• /
• 2010

This paper presents a simulation model based on support vector regression (SVR) for flicker estimation emitted from the wind turbines. For the SVR training, the voltage variation and flicker level are selected as input and output, respectively. Through the off-line training, the relationship between the voltage variation and flicker level is derived. The required amount of data for the flicker measurement is decreased and its proessing time is also reduced. The simulation and experiment results have shown that the flicker estimation is performed accurately.

• Steel and Composite Structures
• /
• v.37 no.2
• /
• pp.193-209
• /
• 2020

In this paper, the efficiency of five Machine Learning (ML) methods consisting of Deep Learning (DL), Support Vector Machine (SVM), Random Forest (RF), Decision Tree (DT), and Gradient Tree Booting (GTB) for regression and classification of the Ultimate Load Factor (ULF) of nonlinear inelastic steel frames is compared. For this purpose, a two-story, a six-story, and a twenty-story space frame are considered. An advanced nonlinear inelastic analysis is carried out for the steel frames to generate datasets for the training of the considered ML methods. In each dataset, the input variables are the geometric features of W-sections and the output variable is the ULF of the frame. The comparison between the five ML methods is made in terms of the mean-squared-error (MSE) for the regression models and the accuracy for the classification models, respectively. Moreover, the ULF distribution curve is calculated for each frame and the strength failure probability is estimated. It is found that the GTB method has the best efficiency in both regression and classification of ULF regardless of the number of training samples and the space frames considered.

• Advances in Computational Design
• /
• v.7 no.2
• /
• pp.113-128
• /
• 2022

Water consumption is strongly affected by numerous factors, such as population, climatic, geographic, and socio-economic factors. Therefore, the implementation of a reliable predictive model of water consumption pattern is challenging task. This study investigates the performance of predictive models based on multi-layer perceptron (MLP), multiple linear regression (MLR), and support vector regression (SVR). To understand the significant factors affecting water consumption, the stepwise regression (SW) procedure is used in MLR to obtain suitable variables. Then, this study also implements three predictive models based on these significant variables (e.g., SWMLR, SWMLP, and SWSVR). Annual data of water consumption in Thailand during 2006 - 2015 were compiled and categorized by provinces and distributors. By comparing the predictive performance of models with all variables, the results demonstrate that the MLP models outperformed the MLR and SVR models. As compared to the models with selected variables, the predictive capability of SWMLP was superior to SWMLR and SWSVR. Therefore, the SWMLP still provided satisfactory results with the minimum number of explanatory variables which in turn reduced the computation time and other resources required while performing the predictive task. It can be concluded that the MLP exhibited the best result and can be utilized as a reliable water demand predictive model for both of all variables and selected variables cases. These findings support important implications and serve as a feasible water consumption predictive model and can be used for water resources management to produce sufficient tap water to meet the demand in each province of Thailand.

• Structural Engineering and Mechanics
• /
• v.25 no.5
• /
• pp.631-636
• /
• 2007

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2016.05a
• /
• pp.285-286
• /
• 2016

We aim to propose the prediction modeling method of ship's position with extracting ship's trajectory model through pattern recognition based on the data that are being collected in VTS centers at real time. Support Vector Machine algorithm was used for data modeling. The optimal parameters are calculated with k-fold cross validation and grid search. We expect that the proposed modeling method could support VTS operators' decision making in case of complex encountering traffic situations.

• Nuclear Engineering and Technology
• /
• v.47 no.1
• /
• pp.74-84
• /
• 2015

If a transient occurs in a nuclear power plant (NPP), operators will try to protect the NPP by estimating the kind of abnormality and mitigating it based on recommended procedures. Similarly, operators take actions based on severe accident management guidelines when there is the possibility of a severe accident occurrence in an NPP. In any such situation, information about the occurrence time of severe accident-related events can be very important to operators to set up severe accident management strategies. Therefore, support systems that can quickly provide this kind of information will be very useful when operators try to manage severe accidents. In this research, the occurrence times of several events that could happen during a severe accident were predicted using support vector machines with short time variations of plant status variables inputs. For the preliminary step, the break location and size of a loss of coolant accident (LOCA) were identified. Training and testing data sets were obtained using the MAAP5 code. The results show that the proposed algorithm can correctly classify the break location of the LOCA and can estimate the break size of the LOCA very accurately. In addition, the occurrence times of severe accident major events were predicted under various severe accident paths, with reasonable error. With these results, it is expected that it will be possible to apply the proposed algorithm to real NPPs because the algorithm uses only the early phase data after the reactor SCRAM, which can be obtained accurately for accident simulations.