• Smart Structures and Systems
• /
• 제14권2호
• /
• pp.145-158
• /
• 2014

The objective of this study is to improve the survivability and reliability of the FBG sensor network in the structural health monitoring (SHM) system. Therefore, a model reconstruction soft computing recognition algorithm based on support vector regression (SVR) is proposed to achieve the high reliability of the FBG sensor network, and the grid search algorithm is used to optimize the parameters of SVR model. Furthermore, in order to demonstrate the effectiveness of the proposed model reconstruction algorithm, a SHM system based on an eight-point fiber Bragg grating (FBG) sensor network is designed to monitor the foreign-object low velocity impact of a CFRP composite plate. Simultaneously, some sensors data are neglected to simulate different kinds of FBG sensor network failure modes, the predicting results are compared with non-reconstruction for the same failure mode. The comparative results indicate that the performance of the model reconstruction recognition algorithm based on SVR has more excellence than that of non-reconstruction, and the model reconstruction algorithm almost keeps the consistent predicting accuracy when no sensor, one sensor and two sensors are invalid in the FBG sensor network, thus the reliability is improved when there are FBG sensors are invalid in the structural health monitoring system.

• Computers and Concrete
• /
• 제32권6호
• /
• pp.577-594
• /
• 2023

Engineered cementitious composites with calcined clay limestone cement (LC³-ECC) as a kind of green, low-carbon and high toughness concrete, has recently received significant investigation. However, the complicated relationship between potential influential factors and LC³-ECC compressive strength makes the prediction of LC³-ECC compressive strength difficult. Regarding this, the machine learning-based prediction models for the compressive strength of LC³-ECC concrete is firstly proposed and developed. Models combine three novel meta-heuristic algorithms (golden jackal optimization algorithm, butterfly optimization algorithm and whale optimization algorithm) with support vector regression (SVR) to improve the accuracy of prediction. A new dataset about LC³-ECC compressive strength was integrated based on 156 data from previous studies and used to develop the SVR-based models. Thirteen potential factors affecting the compressive strength of LC³-ECC were comprehensively considered in the model. The results show all hybrid SVR prediction models can reach the Coefficient of determination (R²) above 0.95 for the testing set and 0.97 for the training set. Radar and Taylor plots also show better overall prediction performance of the hybrid SVR models than several traditional machine learning techniques, which confirms the superiority of the three proposed methods. The successful development of this predictive model can provide scientific guidance for LC³-ECC materials and further apply to such low-carbon, sustainable cement-based materials.

• 지식경영연구
• /
• 제23권1호
• /
• pp.187-201
• /
• 2022

관광객이 작성한 온라인 리뷰는 관광산업의 관리 및 운영에 중요한 정보를 제공한다. 평점은 제품이나 서비스에 대한 정량적인 평가로 간편하지만 관광객의 진실한 태도를 반영하기 어려우며 평점과 리뷰내용에 대한 불일치 문제도 발생하고 있다. 불일치 문제는 잠재고객에게 혼동을 줄 수 있으며 구매의사결정에도 영향을 미칠 수 있다. 본 연구에서는 온라인 리뷰기반의 평점 예측모형을 통해 평점과 리뷰내용의 불일치 문제를 해결하고자 한다. 한국을 방문한 외국인 관광객이 작성한 관광지와 호텔에 대한 리뷰의 감성분석을 통해 평점과 감성의 차이를 비교하고 TF-IDF vectorization과 감성분석 결과로 변수를 선정하였다. 로짓, 인공신경망, SVM(Support Vector Machine)을 적용하여 평점을 분류하고, 인공신경망, SVR(Support Vector Regression)을 통해 평점을 예측하였다. 평점 분류모형과 예측모형 모두 불일치한 리뷰를 제거하고 감성분석을 반영한 모형에서 우수한 성과를 보여주었다. 본 연구에서 제안한 온라인 리뷰 기반의 평점 예측모형은 평점과 리뷰내용에 대한 불일치 문제를 해결하여 신뢰할 수 있는 정보를 제공하였으며 평점이 없는 온라인 리뷰에도 활용할 수 있을 것이다.

• Advances in Computational Design
• /
• 제7권2호
• /
• 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.

• Smart Structures and Systems
• /
• 제14권3호
• /
• pp.397-418
• /
• 2014

In this paper, a simple two-step method for structural vibration-based health monitoring for beam-like structures have been extended to plate-like structures with though-thickness cracks. Crack locations and severities of plate-like structures are detected using a hybrid approach. The interval wavelet transform is employed to extract crack singularity locations from mode shape and support vector regression (SVR) is applied to predict crack serviettes form crack severity detection database (the relationship of natural frequencies and crack serviettes) using several natural frequencies as inputs. Of particular interest is the natural frequencies estimation for cracked plate-like structures using Rayleigh quotient. Only the natural frequencies and mode shapes of intact structures are needed to calculate the natural frequencies of cracked plate-like structures using a simple formula. The crack severity detection database can be easily obtained with this formula. The hybrid method is investigated using numerical simulation and its validity of the usage of interval wavelet transform and SVR are addressed.

• Structural Engineering and Mechanics
• /
• 제73권4호
• /
• pp.455-462
• /
• 2020

The hybrid method using the extended finite element method (XFEM) and the forward Euler approach is widely employed to predict the fatigue life of plate structures. Due to the accuracy of the forward Euler approach is determined by a small step size, the performance of fatigue life prediction of the hybrid method is not agreeable. Instead the forward Euler approach, a prediction method using midpoint method and support vector regression (SVR) is presented to evaluate the stress intensity factors (SIFs) and the fatigue life. Firstly, the XFEM is employed to calculate the SIFs with given crack sizes. Then use the history of SIFs as a function of either number of fatigue life cycles or crack sizes within the current cycle to build a prediction model. Finally, according to the prediction model predict the SIFs at different crack sizes or different cycles. Three numerical cases composed by a homogeneous plate with edge crack, a composite plate with edge crack and center crack are introduced to verify the performance of the proposed method. The results show that the proposed method enables large step sizes without sacrificing accuracy. The method is expected to predict the fatigue life of complex structures.

• Structural Engineering and Mechanics
• /
• 제77권2호
• /
• pp.231-243
• /
• 2021

In this paper, a dynamic response mapping model of the wheel-rail system is established by using the support vector regression (SVR) method, and the hierarchical safety thresholds of the subgrade void are proposed based on the reliability theory. Firstly, the vehicle-track coupling dynamic model considering the subgrade void is constructed. Secondly, the subgrade void area, the subgrade compaction index K30 and the fastener stiffness are selected as random variables, and the mapping model between these three random parameters and the dynamic response of the wheel-rail system is built by using the orthogonal test and the SVR. The sensitivity analysis is carried out by the range analysis method. Finally, the hierarchical safety thresholds for the subgrade void are proposed. The results show that the subgrade void has the most significant influence on the carbody vertical acceleration, the rail vertical displacement, the vertical displacement and the slab tensile stress. From the range analysis, the subgrade void area has the largest effect on the dynamic response of the wheel-rail system, followed by the fastener stiffness and the subgrade compaction index K30. The recommended safety thresholds for the subgrade void of level I, II and III are 4.01㎡, 6.81㎡ and 9.79㎡, respectively.

• 응용통계연구
• /
• 제34권3호
• /
• pp.461-475
• /
• 2021

은닉 마르코프 모델(hidden Markov model, HMM)은 은닉된 상태와 관찰 가능한 결과의 두 가지 요소로 이루어진 통계적 모형으로 확률론적 접근이 가능하고, 다양한 수학적인 구조를 가지고 있어 여러 분야에서 활발하게 사용되고 있다. 특히 금융 분야의 시계열 데이터에 응용되어 다양한 연구가 진행되고 있다. 본 연구는 HMM 이론을 국내 KOSPI200 주가지수와 더불어 NIKKEI225, HSI, S&P500, FTSE100과 같은 해외 주가지수 예측에 적용해 보고자 한다. 또한, 최근 인공지능 분야의 발전으로 인해 주식 가격 예측에 빈번하게 사용되는 서포트 벡터 회귀(support vector regression, SVR) 결과와 어떤 차이가 있는지 비교하여 살펴보고자 한다.

• 한국시뮬레이션학회논문지
• /
• 제18권3호
• /
• pp.103-112
• /
• 2009

원자력발전소에서 센서의 주기적 교정은 안전운전을 위해 꼭 필요하다. 그러나 실제 드리프트가 발생하여 교정을 요하는 센서는 약 2% 미만이다. 또한, 센서의 작동 상태를 매 핵연료 주기마다 수행하는 것은 고장 혹은 드리프트가 발생한 센서를 최대 18개월까지 감지하지 못한 채 운전할 위험이 있다. 원전의 안전운전 및 불필요한 교정을 줄이기 위해 센서의 상시 교정 감시가 필요하다. 이를 위해 주성분 분석과 Support Vector Regression(SVR)을 이용한 PCSVR 알고리즘을 개발하였고, 고리원전 3호기의 출력증발 데이터를 이용하여 검증하였다. 주성분분석은 선형변환을 통한 입력공간의 축소 및 노이즈 제거 효과를 나타내며, AASVR은 해석학적 및 기계학적 모델로 모델링하기 힘든 복잡계를 쉽게 나타낼 수 있는 장점이 있다. SVR의 세가지 파라미터는 반응표면분석법에 의해 최적화하였다. 센서의 고장탐지를 위해 모델 출력의 잔차를 슈하르트 관리도, EWMA, CUSUM 및 일반화우도비검정(GLRT)을 통해 그 결과를 비교하였다. 미세한 드리프트에 대해 CUSUM과 GLRT가 우수한 결과를 보였다. 개발된 알고리즘은 수출형 원전 APR1000 설계시 적용가능 할 것으로 판단된다.

• Steel and Composite Structures
• /
• 제51권5호
• /
• pp.509-527
• /
• 2024

Bonding carbon fiber-reinforced polymer (CFRP) laminates have been extensively employed in the restoration of steel constructions. In addition to the mechanical properties of the CFRP, the bond strength (P_U) between the CFRP and steel is often important in the eventual strengthened performance. Nonetheless, the bond behavior of the CFRP-steel (CS) interface is exceedingly complicated, with multiple failure causes, giving the P_U challenging to forecast, and the CFRP-enhanced steel structure is unsteady. In just this case, appropriate methods were established by hybridized Random Forests (RF) and support vector regression (SVR) approaches on assembled CS single-shear experiment data to foresee the P_U of CS, in which a recently established optimization algorithm named Aquila optimizer (AO) was used to tune the RF and SVR hyperparameters. In summary, the practical novelty of the article lies in its development of a reliable and efficient method for predicting bond strength at the CS interface, which has significant implications for structural rehabilitation, design optimization, risk mitigation, cost savings, and decision support in engineering practice. Moreover, the Fourier Amplitude Sensitivity Test was performed to depict each parameter's impact on the target. The order of parameter importance was t_c> L_c > E_A > t_A > E_c > b_c > f_c > f_A from largest to smallest by 0.9345 > 0.8562 > 0.79354 > 0.7289 > 0.6531 > 0.5718 > 0.4307 > 0.3657. In three training, testing, and all data phases, the superiority of AO - RF with respect to AO - SVR and MARS was obvious. In the training stage, the values of R² and VAF were slightly similar with a tiny superiority of AO - RF compared to AO - SVR with R² equal to 0.9977 and VAF equal to 99.772, but large differences with results of MARS.