• Analytical Science and Technology
• /
• v.28 no.4
• /
• pp.270-277
• /
• 2015

Boiling point (BP) is one of the most fundamental physicochemical properties of organic compounds to characterize and identify the thermal characteristics of target compounds. Previously developed QSPR equations, however, still had some limitation for the specific compounds, like high-energy molecules, mainly because of the lack of experimental data and less coverage. A large BP dataset of 5,923 solid organic compounds was finally secured in this study, after dedicated pre-filtration of experimental data from different sources, mostly consisting of compounds not only from common organic molecules but also from some specially used molecules, and those dataset was used to build the new BP prediction model. Various machine learning methods were performed for newly collected data based on meaningful 2D descriptor set. Results of combined check showed acceptable validity and robustness of our models, and consensus approaches of each model were also performed. Applicability domain of BP prediction model was shown based on descriptor of training set.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.12 no.4
• /
• pp.291-299
• /
• 2007

To operate a permanent magnet synchronous motor (PMSM) at a maximum torque per ampere (MTPA) operation point, the exact values of machine parameters such as inductances and back-EMF constant, which are sensitive to motor phase currents and temperature respectively, should be blown. An adaptive estimation method for on-line estimation of the machine parameters is not suitable for practical applications since it has difficulties in estimating exact values and requires complex mathematical calculations. The purpose of this paper is to present a simple MTPA operation point tracking control strategy for vector controlled PMSM drives with slow dynamic loads. The proposed method searches MTPA operation points by modulating current phase angle and observing the variation in command power. The current angle modulation strategy is designed to sense the effect of load variations in the command power. Therefore, the proposed method can track the MTPA operation points of the PMSM regardless of load variations. Computer simulation and experimental study is also presented to show the effectiveness of the proposed method.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.15 no.2
• /
• pp.142-149
• /
• 2010

Interior Permanent Magnet Synchronous Motor(IPMSM) have gained an increasing popularity in recent years for a variety of industrial applications, because of their high power density, high efficiency and possibility of flux weakening operation. Because the efficiency of IPMSM is one of the important performance characteristic, the Maximum Torque Per Ampere(MTPA) operating method has been indispensible. In theory, MTPA operating point can be calculated using the exact values of the machine parameters. However, the values of the IPMSM parameters are known to vary widely according to the operating condition. Therefore, to operate the IPMSM in the MTPA operating point, the machine parameters should be estimated in real-time. In this paper, the new MTPA operating method based on the signal injection is presented. By injecting the high frequency current signal, the MTPA operating criteria can be calculated by measuring the input power to IPMSM. The proposed method can find the MTPA operating point with simple signal processing regardless of the parameter variation.

• Steel and Composite Structures
• /
• v.45 no.2
• /
• pp.159-173
• /
• 2022

This paper proposes a hybrid machine-learning model, which is called DANN-IP, that combines a deep artificial neural network (DANN) and an interior-point (IP) algorithm in order to improve the prediction capacity on the patch loading resistance of steel plate girders. For this purpose, 394 steel plate girders that were subjected to patch loading were tested in order to construct the DANN-IP model. Firstly, several DANN models were developed in order to establish the relationship between the patch loading resistance and the web panel length, the web height, the web thickness, the flange width, the flange thickness, the applied load length, the web yield strength, and the flange yield strength of steel plate girders. Accordingly, the best DANN model was chosen based on three performance indices, which included the R^2, RMSE, and a20-index. The IP algorithm was then adopted to optimize the weights and biases of the DANN model in order to establish the hybrid DANN-IP model. The results obtained from the proposed DANN-IP model were compared with of the results from the DANN model and the existing empirical formulas. The comparison showed that the proposed DANN-IP model achieved the best accuracy with an R^2 of 0.996, an RMSE of 23.260 kN, and an a20-index of 0.891. Finally, a Graphical User Interface (GUI) tool was developed in order to effectively use the proposed DANN-IP model for practical applications.

• Journal of Institute of Control, Robotics and Systems
• /
• v.22 no.3
• /
• pp.241-246
• /
• 2016

Nowadays many people have an interest in facial expression and the behavior of people. These are human-robot interaction (HRI) researchers utilize digital image processing, pattern recognition and machine learning for their studies. Facial feature point detector algorithms are very important for face recognition, gaze tracking, expression, and emotion recognition. In this paper, a cascade facial feature point detector is used for finding facial feature points such as the eyes, nose and mouth. However, the detector has difficulty extracting the feature points from several images, because images have different conditions such as size, color, brightness, etc. Therefore, in this paper, we propose an algorithm using a modified cascade facial feature point detector using a convolutional neural network. The structure of the convolution neural network is based on LeNet-5 of Yann LeCun. For input data of the convolutional neural network, outputs from a cascade facial feature point detector that have color and gray images were used. The images were resized to $32{\times}32$. In addition, the gray images were made into the YUV format. The gray and color images are the basis for the convolution neural network. Then, we classified about 1,200 testing images that show subjects. This research found that the proposed method is more accurate than a cascade facial feature point detector, because the algorithm provides modified results from the cascade facial feature point detector.

• Computers and Concrete
• /
• v.32 no.5
• /
• pp.465-474
• /
• 2023

The point load test (PLT) is a widely-used alternative method in the field to determine the uniaxial compressive strength due to its simple testing machine and procedure. The point load test index can estimate the uniaxial compressive strength through conversion factors based on the rock types. However, the mechanism correlating these two parameters and the influence of the mechanical properties on PLT results are still not well understood. This study proposed a theoretical model to understand the mechanism of PLT serving as an alternative to the UCS test based on laboratory observation and literature survey. This model found that the point load test is a self-confined compression test. There is a compressive ellipsoid near the loading axis, whose dilation forms a tensile ring that provides confinement on this ellipsoid. The peak load of a point load test is linearly positive correlated to the tensile strength and negatively correlated to the Poisson ratio. The model was then verified using numerical and experimental approaches. In numerical verification, the PLT discs were simulated using flat-joint BPM of PFC3D to model the force distribution, crack propagation and BPM properties' effect with calibrated micro-parameters from laboratory UCS test and point load test of Berea sandstones. It further verified the mechanism experimentally by conducting a uniaxial compressive test, Brazilian test, and point load test on four different rocks. The findings from this study can explain the mechanism and improve the understanding of point load in determining uniaxial compressive strength.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.28 no.2
• /
• pp.370-376
• /
• 2022

Smart ships related studies are being conducted in various fields owing to the development of technology, and an engine room patrol robot that can patrol the unmanned engine room is one such study. A patrol robot moves around the engine room based on the information learned through artificial intelligence and checks the machine normality and occurrence of abnormalities such as water leakage, oil leakage, and fire. Study on engine room patrol robots is mainly conducted on machine detection using artificial intelligence, however study on movement and control is insufficient. This causes a problem in that even if a patrol robot detects an object, there is no way to move to the detected object. To secure maneuverability to quickly identify the presence of abnormality in the engine room, this study experimented with whether a patrol robot can determine the shortest path by applying the A* algorithm. Data were obtained by driving a small car equipped with LiDAR in the ship engine room and creating a map by mapping the obtained data with SLAM(Simultaneous Localization And Mapping). The starting point and arrival point of the patrol robot were set on the map, and the A* algorithm was applied to determine whether the shortest path from the starting point to the arrival point was found. Simulation confirmed that the shortest route was well searched while avoiding obstacles from the starting point to the arrival point on the map. Applying this to the engine room patrol robot is believed to help improve ship safety.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.17 no.8
• /
• pp.2001-2011
• /
• 1993

This study deals with the machining of sculptured surfaces on 5-axis CNC milling machine with end mill cutter. The study (I) has the following contents. In 5-axis CNC milling, CL-data consist of CC-data and cutter axis direction vector at the CC-point. Thus, in machining of the sculptured surface on 5-axis CNC milling machine, determination of the direction vector of the milling cutter is very important. The direction vector is obtained by the fact that bottom plane of the milling cutter must not interfere with the free-form surface being machined. The interference is checked by the z-map method which can be applied in all geometric types of the sculptured surfaces. After generating NC part programs from 5-axis post-processing algorithms, sculptured surfaces were machined with 5-axis CNC milling machine (CINCINNATI MILACRON, 20V-80). From these machining tests, it was shown that the machining of the free-form surfaces on 5-axis CNC milling machine with the end mill has smaller cusp heights and shorter cutting time than on 3-axis CNC milling machine with the ball-end mill. Thus, 5-axis CNC end milling was effective machining method for sculptured surfaces. The study (II) deals with the prediction of cusp height and the determination of tool path interval for the 5-axis machining of sculptured surfaces on the basis of study(I).

• Archives of design research
• /
• no.18
• /
• pp.177-186
• /
• 1996

With the change of patterns and the environment of the national standard of living the prohlem of environmental pollution became increasingly serious. Because of the enormous increase of various kinds of used and (after utilization) useless articles, efforts to save resources as well as the environment and the promotion of reprated utilization and recycling are inavoidable. The recognition of an environmental an health problem, and the desire for nonpollution created a desire for environment-friendly products in order to avoid an environmental consumptionism. Drink vending machines making use of vessels only once are closely related to the environmental problem. It is therefore necessary to develop an ecologically designed vending machine. In this study the backgrounds and concepts of green design, classification, construction and the environment of a drink vending machine arc analyzed. From this st1.rting-point a concept for the design of a drink vending machine is developed by two concepts : Type A (seperated-gathering type) and Type B (recycling type). Then three defferent types of vending-machines arc introduced a wall -adherable type, a center est1.blishable type and a desk top type. The conclusion of the text is threefold. There are needs for an ecological design of vending machines, ergonomIc considerations and a harmonization of the styldapperarance) of the machine and its circumferences.

• Journal of Convergence for Information Technology
• /
• v.8 no.1
• /
• pp.201-206
• /
• 2018

Recently, as technologies for realizing artificial intelligence have become more common, machine learning is widely used. Machine learning provides insight into collecting large amounts of data, batch processing, and taking final action, but the effects of the work are not immediately integrated into the learning process. In this paper proposed an adaptive learning model to improve the performance of real-time stream analysis as a big business issue. Adaptive learning generates the ensemble by adapting to the complexity of the data set, and the algorithm uses the data needed to determine the optimal data point to sample. In an experiment for six standard data sets, the adaptive learning model outperformed the simple machine learning model for classification at the learning time and accuracy. In particular, the support vector machine showed excellent performance at the end of all ensembles. Adaptive learning is expected to be applicable to a wide range of problems that need to be adaptively updated in the inference of changes in various parameters over time.