• Journal of IKEEE
• /
• v.23 no.1
• /
• pp.127-133
• /
• 2019

Electric power demand forecasting is one of the important research areas for future smart grid introduction. However, It is difficult to predict because it is affected by many external factors. Traditional methods of forecasting power demand have been limited in making accurate prediction because they use raw power data. In this paper, a probability-based CRBM is proposed to solve the problem of electric power demand prediction using raw power data. The stochastic model is suitable to capture the probabilistic characteristics of electric power data. In order to compare the mid-term power demand forecasting performance of the proposed model, we compared the performance with Recurrent Neural Network(RNN). Performance comparison using electric power data provided by the University of Massachusetts showed that the proposed algorithm results in better performance in mid-term energy demand forecasting.

• Design & Manufacturing
• /
• v.15 no.1
• /
• pp.48-56
• /
• 2021

The quality of products produced by injection molding process is greatly influenced by the process variables set on the injection molding machine during manufacturing. It is very difficult to predict the quality of injection molded product considering the stochastic nature of manufacturing process, because the process variables complexly affect the quality of the injection molded product. In the present study we predicted the quality of injection molded product using Artificial Neural Network (ANN) method specifically from Multiple Input Single Output (MISO) and Multiple Input Multiple Output (MIMO) perspectives. In order to train the ANN model a systematic plan was prepared based on a combination of orthogonal sampling and random sampling methods to represent various and robust patterns with small number of experiments. According to the plan the injection molding experiments were conducted to generate data that was separated into training, validation and test data groups to optimize the parameters of the ANN model and evaluate predicting performance of 4 structures (MISO1-2, MIMO1-2). Based on the predicting performance test, it was confirmed that as the number of output variables were decreased, the predicting performance was improved. The results indicated that it is effective to use single output model when we need to predict the quality of injection molded product with high accuracy.

• Smart Structures and Systems
• /
• v.31 no.4
• /
• pp.421-436
• /
• 2023

This paper presents a vibration displacement measurement and damage identification method for a space truss structure from its vibration videos. Features from Accelerated Segment Test (FAST) algorithm is combined with adaptive threshold strategy to detect the feature points of high quality within the Region of Interest (ROI), around each node of the truss structure. Then these points are tracked by Kanade-Lucas-Tomasi (KLT) algorithm along the video frame sequences to obtain the vibration displacement time histories. For some cases with the image plane not parallel to the truss structural plane, the scale factors cannot be applied directly. Therefore, these videos are processed with homography transformation. After scale factor adaptation, tracking results are expressed in physical units and compared with ground truth data. The main operational frequencies and the corresponding mode shapes are identified by using Subspace Stochastic Identification (SSI) from the obtained vibration displacement responses and compared with ground truth data. Structural damages are quantified by elemental stiffness reductions. A Bayesian inference-based objective function is constructed based on natural frequencies to identify the damage by model updating. The Success-History based Adaptive Differential Evolution with Linear Population Size Reduction (L-SHADE) is applied to minimise the objective function by tuning the damage parameter of each element. The locations and severities of damage in each case are then identified. The accuracy and effectiveness are verified by comparison of the identified results with the ground truth data.

• International conference on construction engineering and project management
• /
• 2009.05a
• /
• pp.700-706
• /
• 2009

Bridges are vital components of any road network which demand crucial and timely decision-making for Maintenance, Repair and Rehabilitation (MR&R) activities. Bridge Management Systems (BMSs) as a decision support system (DSS), have been developed since the early 1990's to assist in the management of a large bridge network. Historical condition ratings obtained from biennial bridge inspections are major resources for predicting future bridge deteriorations via BMSs. Available historical condition ratings in most bridge agencies, however, are very limited, and thus posing a major barrier for obtaining reliable future structural performances. To alleviate this problem, the verified Backward Prediction Model (BPM) technique has been developed to help generate missing historical condition ratings. This is achieved through establishing the correlation between known condition ratings and such non-bridge factors as climate and environmental conditions, traffic volumes and population growth. Such correlations can then be used to obtain the bridge condition ratings of the missing years. With the help of these generated datasets, the currently available bridge deterioration model can be utilized to more reliably forecast future bridge conditions. In this paper, the prediction accuracy based on 4 and 9 BPM-generated historical condition ratings as input data are compared, using deterministic and stochastic bridge deterioration models. The comparison outcomes indicate that the prediction error decreases as more historical condition ratings obtained. This implies that the BPM can be utilised to generate unavailable historical data, which is crucial for bridge deterioration models to achieve more accurate prediction results. Nevertheless, there are considerable limitations in the existing bridge deterioration models. Thus, further research is essential to improve the prediction accuracy of bridge deterioration models.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.25 no.6
• /
• pp.447-458
• /
• 2023

In this study, a comparison model considering the stochastic characteristics of the load and member resistance of the shield tunnel segment lining as well as the variability of the boundary condition was selected and reliability analysis was performed, and the adequacy of the limit state design was analyzed by calculating the probability of failure and reviewing the structural safety. For the analysis considering the probability characteristics of these ground constants, the ground spring coefficient was considered as the mean value by calculating the quantitative value by applying the Muirwood formula, and the coefficient of variation was selected based on the existing research data to review the models according to the change of ground boundary conditions. Through the structural analysis of these models and the reliability analysis using MCS technique, the failure probability and reliability index were calculated to examine the changes in the failure probability due to changes in ground boundary conditions.

• Wind and Structures
• /
• v.10 no.4
• /
• pp.367-382
• /
• 2007

A nonlinear numerical method was developed to assess the stability of suspension bridge catwalks under a wind load. A section model wind tunnel test was used to obtain a catwalk's aerostatic coefficients, from which the displacement-dependent wind loads were subsequently derived. The stability of a suspension bridge catwalk was analyzed on the basis of the geometric nonlinear behavior of the structure. In addition, a full model test was conducted on the catwalk, which spanned 960 m. A comparison of the displacement values between the test and the numerical simulation shows that a numerical method based on a section model test can be used to effectively and accurately evaluate the stability of a catwalk. A case study features the stability of the catwalk of the Runyang Yangtze suspension bridge, the main span of which is 1490 m. Wind can generally attack the structure from any direction. Whenever the wind comes at a yaw angle, there are six wind load components that act on the catwalk. If the yaw angle is equal to zero, the wind is normal to the catwalk (called normal wind) and the six load components are reduced to three components. Three aerostatic coefficients of the catwalk can be obtained through a section model test with traditional test equipment. However, six aerostatic coefficients of the catwalk must be acquired with the aid of special section model test equipment. A nonlinear numerical method was used study the stability of a catwalk under a yaw wind, while taking into account the six components of the displacement-dependent wind load and the geometric nonlinearity of the catwalk. The results show that when wind attacks with a slight yaw angle, the critical velocity that induces static instability of the catwalk may be lower than the critical velocity of normal wind. However, as the yaw angle of the wind becomes larger, the critical velocity increases. In the atmospheric boundary layer, the wind is turbulent and the velocity history is a random time history. The effects of turbulent wind on the stability of a catwalk are also assessed. The wind velocity fields are regarded as stationary Gaussian stochastic processes, which can be simulated by a spectral representation method. A nonlinear finite-element model set forepart and the Newmark integration method was used to calculate the wind-induced buffeting responses. The results confirm that the turbulent character of wind has little influence on the stability of the catwalk.

• Journal of Educational Research in Mathematics
• /
• v.22 no.3
• /
• pp.371-386
• /
• 2012

This research assumes two facts; One is that the mathematics curriculum reform of Korea in 2007 would have been better if it had been a revise instead of deletion and the other is that every school curriculum should be of help for the sound enhancement of all 6 types of logical concepts that appears in the Piaget's theory of cognitive development. What our mathematics curriculum has introduced as a correlation is not the one of the 6 logical concepts that Piaget had thought in his theory of cognitive development. In order to see the reason of that difference, we check the difference of jargons among the academic denominations, such as Pedagogy, Psychology and Statistics through their college textbooks. Because we suppose that the mismatch of 'Piaget's vs Curriculum's correlation' is due to the mis-communication among scholars of different academic denominations. With what we learned via the above analytical study leaned on an abduction and to get some idea on them for the potential future construction of school Statistics curriculum when it should be returned, which we believe so, we observe two foreign highschool mathematics textbooks briefly. As a result of the study, we found that the concept of correlation in Pedagogy contain all kinds of relation while it was stingy in Statistics. Here we report a main result; A careful discretion among similar concepts of correlation, such as linear relationship(correlation), stochastic change along conditions(dependence), central comparison(other relation) are needed for the potential future curriculum. And if new curriculum contains the linear correlation then we strongly recommend to involve the regression line to connect it with the linear function chapter.

• Korean Journal of Biological Psychiatry
• /
• v.27 no.1
• /
• pp.18-26
• /
• 2020

Objectives The aim was to find effective vectorization and classification models to predict a psychiatric diagnosis from text-based medical records. Methods Electronic medical records (n = 494) of present illness were collected retrospectively in inpatient admission notes with three diagnoses of major depressive disorder, type 1 bipolar disorder, and schizophrenia. Data were split into 400 training data and 94 independent validation data. Data were vectorized by two different models such as term frequency-inverse document frequency (TF-IDF) and Doc2vec. Machine learning models for classification including stochastic gradient descent, logistic regression, support vector classification, and deep learning (DL) were applied to predict three psychiatric diagnoses. Five-fold cross-validation was used to find an effective model. Metrics such as accuracy, precision, recall, and F1-score were measured for comparison between the models. Results Five-fold cross-validation in training data showed DL model with Doc2vec was the most effective model to predict the diagnosis (accuracy = 0.87, F1-score = 0.87). However, these metrics have been reduced in independent test data set with final working DL models (accuracy = 0.79, F1-score = 0.79), while the model of logistic regression and support vector machine with Doc2vec showed slightly better performance (accuracy = 0.80, F1-score = 0.80) than the DL models with Doc2vec and others with TF-IDF. Conclusions The current results suggest that the vectorization may have more impact on the performance of classification than the machine learning model. However, data set had a number of limitations including small sample size, imbalance among the category, and its generalizability. With this regard, the need for research with multi-sites and large samples is suggested to improve the machine learning models.

• Journal of Korean Society of Transportation
• /
• v.23 no.1
• /
• pp.67-80
• /
• 2005

In this paper, the primary objective of the research are to review the methods currently avaliable for estimating the delay incurred by vehicles at signalized intersections. The paper compares the delay estimates from a deterministic queueing model, a model based on shock wave theory , the steady-state Webster model, the queue-based models defined in the 1994 and 2001 version of the High way Capacity Manual, in addition to the delays estimated from the TRANSYT-7F macroscopic simulation and NETSIM microscopic simulation. More especially, this paper is to compare the delay estimates obtained using macroscopic and microscopic simulation tools against state-of-the practice analytical models that are derived from deterministic queueing and shock wave analysis theory. The results of the comparisons indicate that all delay models produce relatively similar results for signalized intersections with low traffic demand, but that increasing differences occur as the traffic demand approaches saturation. In particular, when the TRANSYT-7F and NETSIM are compared, it is highly differences as approach for traffic condition to over-saturation. Also, the NETSIM microscopic simulation is the lowest estimates among the various models.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.49 no.5
• /
• pp.373-381
• /
• 2021

This paper details the result of flight experiment that examines performance of roll angle estimation algorithm of slowly rolling munition taking time delay of measurement into account when measurement comes in delayed fashion. As the measurement is passed through low pass filter for numerical stabilization and de-noising purpose which induces time delay, we design augmented state Kalman filter that incorporates distribution models of stochastic delay over time. Flight experiment was conducted to verify the algorithm at around 250m high AGL(Above Ground Level) conveying velocity of 28m/s from fixed-wing mother plane to the munition. Munition was made spun with respect to its roll axis using internal reaction wheel afterward. Numerical comparison of proposing method's roll estimation performance with that of commercial aerospace graded GPS/INS shows that proposed filter design can effectively compensate time delay of measurement.