• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.26 no.3D
• /
• pp.461-467
• /
• 2006

To detect individual vehicular speed, double loop detection technique has been widely used. This paper investigates four methodologies to measure individual speed using only a single loop sensor in a traveling lane. Two methods developed earlier include estimating the speed by means of (Case 1) the slop of inductance wave form generated by the sensor and (Case 2) the average vehicle lengths. Two other methods are newly developed through this study, which are estimations by measuring (Case 3) the mean of wheelbases using the sensor installed traversal to the traveling lane and (Case 4) the mean of wheel tracks by the sensor installed diagonally to the traveling lane. These four methodologies were field-tested and their accuracy of speed output was compared statistically. This study used Equality Coefficient and Mean Absolute Percentage Error for the assessment. It was found that the method (Case 1) was best accurate, followed by method (Case 4), (Case 2), and (Case 3).

• Journal of Drive and Control
• /
• v.21 no.1
• /
• pp.22-30
• /
• 2024

Gearboxes designed for reciprocating motion operating mechanisms operate under conditions where both the load and speed undergo continuous variations. When conducting durability tests on gearboxes designed for such applications, operating the target gearbox under conditions similar to the intended usage is essential. The gearbox must be operated for the required number of cycles to validate its durability under conditions mirroring its intended usage. This study devised an accelerated test method for gearboxes, which reduces operating angles and operational strokes. The reliability of the accelerated test was verified by comparing the stresses imposed on the gears under general and acceleration conditions through multi-body dynamic simulations. The results confirmed that the maximum contact stress levels under normal and accelerated conditions were within a 0.1% error range, indicating a minimal difference in the gear damage rates. However, a difference in the maximum contact stress results between the normal and accelerated conditions was observed when inertial forces acted on the output shaft due to the operational acceleration of the gearbox. Therefore, when conducting this acceleration test, caution should be exercised to ensure that the operational load on the gearbox, which affects inertia, does not significantly deviate from the conditions observed under normal operating conditions.

• Journal of Intelligence and Information Systems
• /
• v.17 no.4
• /
• pp.157-173
• /
• 2011

As the Internet use explodes recently, the malicious attacks and hacking for a system connected to network occur frequently. This means the fatal damage can be caused by these intrusions in the government agency, public office, and company operating various systems. For such reasons, there are growing interests and demand about the intrusion detection systems (IDS)-the security systems for detecting, identifying and responding to unauthorized or abnormal activities appropriately. The intrusion detection models that have been applied in conventional IDS are generally designed by modeling the experts' implicit knowledge on the network intrusions or the hackers' abnormal behaviors. These kinds of intrusion detection models perform well under the normal situations. However, they show poor performance when they meet a new or unknown pattern of the network attacks. For this reason, several recent studies try to adopt various artificial intelligence techniques, which can proactively respond to the unknown threats. Especially, artificial neural networks (ANNs) have popularly been applied in the prior studies because of its superior prediction accuracy. However, ANNs have some intrinsic limitations such as the risk of overfitting, the requirement of the large sample size, and the lack of understanding the prediction process (i.e. black box theory). As a result, the most recent studies on IDS have started to adopt support vector machine (SVM), the classification technique that is more stable and powerful compared to ANNs. SVM is known as a relatively high predictive power and generalization capability. Under this background, this study proposes a novel intelligent intrusion detection model that uses SVM as the classification model in order to improve the predictive ability of IDS. Also, our model is designed to consider the asymmetric error cost by optimizing the classification threshold. Generally, there are two common forms of errors in intrusion detection. The first error type is the False-Positive Error (FPE). In the case of FPE, the wrong judgment on it may result in the unnecessary fixation. The second error type is the False-Negative Error (FNE) that mainly misjudges the malware of the program as normal. Compared to FPE, FNE is more fatal. Thus, when considering total cost of misclassification in IDS, it is more reasonable to assign heavier weights on FNE rather than FPE. Therefore, we designed our proposed intrusion detection model to optimize the classification threshold in order to minimize the total misclassification cost. In this case, conventional SVM cannot be applied because it is designed to generate discrete output (i.e. a class). To resolve this problem, we used the revised SVM technique proposed by Platt(2000), which is able to generate the probability estimate. To validate the practical applicability of our model, we applied it to the real-world dataset for network intrusion detection. The experimental dataset was collected from the IDS sensor of an official institution in Korea from January to June 2010. We collected 15,000 log data in total, and selected 1,000 samples from them by using random sampling method. In addition, the SVM model was compared with the logistic regression (LOGIT), decision trees (DT), and ANN to confirm the superiority of the proposed model. LOGIT and DT was experimented using PASW Statistics v18.0, and ANN was experimented using Neuroshell 4.0. For SVM, LIBSVM v2.90-a freeware for training SVM classifier-was used. Empirical results showed that our proposed model based on SVM outperformed all the other comparative models in detecting network intrusions from the accuracy perspective. They also showed that our model reduced the total misclassification cost compared to the ANN-based intrusion detection model. As a result, it is expected that the intrusion detection model proposed in this paper would not only enhance the performance of IDS, but also lead to better management of FNE.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.3 no.2
• /
• pp.18-35
• /
• 1993

The water level control system of the steam generator in a pressurized water reactor and its control problems are analysed. In this work the stable control strategy during the low power operation and transient states is studied. To solve the problem, a fuzzy logic control method is applied as a basic algorithm of the controller. The control algorithm is based on the operator's knowledges and the experiences of manual operation for water level control at the compact nuclear simulator set up in Korea Atomic Energy Research Institute. From a viewpoint of the system realization, the control variables and rules are established considering simpler tuning and the input-output relation. The control strategy includes the dynamic tuning method and employs a substitutional information using the bypass valve opening instead of incorrectly measured signal at the low flow rate as the fuzzy variable of the flow rate during the pressure control mode of the steam generator. It also involves the switching algorithm between the control valves to suppress the perturbation of water level. The simulation results show that both of the fine control action at the small level error and the quick response at the large level error can be obtained and that the performance of the controller is improved.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.40 no.12
• /
• pp.1086-1092
• /
• 2012

It has been studied for DGPS/INS(Differential Global Positioning System/Inertial Navigation System) to offer the more precise and reliable navigation data with the aviation industry development. The flight performance evaluation of navigation system is very significant because the reliability of navigation data directly affect the safety of aircraft. Especially, the high-level navigation system, as DGPS/INS, need more precise flight performance evaluation method. The performance analysis is performed by comparing between the DGPS/INS navigation data and reference trajectory which is more precise than DGPS/INS. The GPS receiver, which is capable of post-processed CDGPS(Carrier-phase DGPS) method, can be used as reference system. Generally, the DGPS/INS is estimated the CG(Center of Gravity) point of aircraft while the reference system is output the position of GPS antenna which is mounted on the outside of aircraft. For this reason, estimated error between DGPS/INS and reference system will include the error due to lever arm. In order to more precise performance evaluation, it is needed to compensate the lever arm. This paper presents procedure and result of flight test which includes lever arm compensation in order to verify reliability and performance of DGPS/INS more precisely.

• Journal of Korean Society of Forest Science
• /
• v.111 no.2
• /
• pp.287-301
• /
• 2022

This study aims to propose a crane-tip control method to intuitively control the end-effector vertically or horizontally for improving the crane work efficiency and to confirm the control performance. To verify the control performance based on experimental variables, a laboratory-scale crane was manufactured using an electric cylinder. Through a forward and reverse kinematics analysis, the crane was configured to output the position coordinates of the current crane-tip and the joint angle at each target point. Furthermore, a method of generating waypoints was used, and a dead band using lateral boundary offset (LBO) was set. Appropriate parameters were selected using bang-bang control, which confirmed that the number of waypoints and LBO radius were associated with positioning error, and the cylinder speed was related to the lead time. With increased number of waypoints and decreased LBO radius, the positioning error and the lead time also decreased as the cylinder speed decreased. Using the proportional control, when the cylinder velocity was changed at every control cycle, the lead time was greatly reduced; however, the actual control pattern was controlled by repeating over and undershoot in a large range. Therefore, proportional control was performed by additionally applying velocity gain that can relatively change the speed of each cylinder. Since the control performed with in a range of 10 mm, it was verified th at th e crane-tip control can be ach ieved with only th e proportional control to which the velocity gain was applied in a control cycle of 20 ms.

• The Journal of Korean Society for Radiation Therapy
• /
• v.31 no.1
• /
• pp.17-24
• /
• 2019

Purpose: The present study aims to assess the level of coherency and the accuracy of Point dose of the Isocenter of VERO, a linear accelerator developed for the purpose of the Stereotactic Body Radiation Therapy(SBRT). Materials and Method: The study was conducted randomly with 10 treatment plans among SBRT patients in Kyungpook National University Chilgok Hospital, using VERO, a linear accelerator between June and December, 2018. In order to assess the equipment's power stability level, we measured the output constancy by using PTW-LinaCheck, an output detector. We also attempted to measure the level of accuracy of the equipment's Laser, kV(Kilo Voltage) imaging System, and MV(Mega Voltage) Beam by using Tofu Phantom(BrainLab, Germany) to assess the accuracy level of geometrical Isocenter. We conducted a comparative analysis to assess the accuracy level of the dose by using an acrylic Phantom($30{\times}30{\times}20cm$), a calibrated ion chamber CC-01(IBA Dosimetry), and an Electrometer(IBA, Dosimetry). Results: The output uniformity of VERO was calculated to be 0.66 %. As for geometrical Isocenter accuracy, we analyzed the error values of ball Isocenter of inner Phantom, and the results showed a maximum of 0.4 mm, a minimum of 0.0 mm, and an average of 0.28 mm on X-axis, and a maximum of -0.4 mm, a minimum of 0.0 mm, and an average of -0.24 mm on Y-axis. A comparison and evaluation of the treatment plan dose with the actual measured dose resulted in a maximum of 0.97 % and a minimum of 0.08 %. Conclusion: The equipment's average output dose was calculated to be 0.66 %, meeting the ${\pm}3%$ tolerance, which was considered as a much uniform fashion. As for the accuracy assessment of the geometric Isocenter, the results met the recommended criteria of ${\pm}1mm$ tolerance, affirming a high level of reproducibility of the patient's posture. The difference between the treatment plan dose and the actual measurement dose was calculated to be 0.52 % on average, significantly less than the 3 % tolerance, confirming that it obtained predicted does. The current study suggested that VERO equipment is suitable for SBRT, and would result in notable therapeutic effect.

• Journal of the Institute of Convergence Signal Processing
• /
• v.12 no.4
• /
• pp.333-340
• /
• 2011

KJJVC(Korea-Japan Joint VLBI Correlator) was installed at the KJCC(Korea-Japan Correlation Center) and has been operated by KASI(Korea Astronomy and Space Science Institute) from 2009. KJNC is able to correlate the VLBI observed data through KVN(Korean VLBI Network), VERA(VLBI Exploration of Radio Astrometry), and JVN(Japanese VLBI Network) and its joint network array. And it is used exclusively as computer in order to process the observed data for the scientific purpose KJJVC used the VSI(VLBI Standard Interface) as the VLBI international standard at the data input-output specification between each component. Especially, for correlating the observed data, the data is transmitted with 1024Mbps speed between Mark5B high-speed playback and RVDB(Raw VLBI Data Buffer). The EMI(Electromagnetic lnterference), which is occurred by data transmission with high-speed, cause the data loss and the loss occurrence is frequently often for long transmission cable. Finally it will be caused the data recognition error by decreasing the voltage level of digital data signal. In this paper, in order to minimize the data loss by measuring the EMI noise level in transmission of the VSI specification, the 3 methods such as 1) RC filtering method, 2) lmpedance matching using Microstrip line, and 3) Signal buffering method using Differential line driver, were proposed. To verify the effectiveness of each proposed method, the performance evaluation was conducted by implementing and simulations for each method. Each proposed method was effectively confirmed as the high-speed data transmission of the VSI specification.

• Journal of Intelligence and Information Systems
• /
• v.22 no.2
• /
• pp.127-142
• /
• 2016

Deep learning model is a kind of neural networks that allows multiple hidden layers. There are various deep learning architectures such as convolutional neural networks, deep belief networks and recurrent neural networks. Those have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition and bioinformatics where they have been shown to produce state-of-the-art results on various tasks. Among those architectures, convolutional neural networks and recurrent neural networks are classified as the supervised learning model. And in recent years, those supervised learning models have gained more popularity than unsupervised learning models such as deep belief networks, because supervised learning models have shown fashionable applications in such fields mentioned above. Deep learning models can be trained with backpropagation algorithm. Backpropagation is an abbreviation for "backward propagation of errors" and a common method of training artificial neural networks used in conjunction with an optimization method such as gradient descent. The method calculates the gradient of an error function with respect to all the weights in the network. The gradient is fed to the optimization method which in turn uses it to update the weights, in an attempt to minimize the error function. Convolutional neural networks use a special architecture which is particularly well-adapted to classify images. Using this architecture makes convolutional networks fast to train. This, in turn, helps us train deep, muti-layer networks, which are very good at classifying images. These days, deep convolutional networks are used in most neural networks for image recognition. Convolutional neural networks use three basic ideas: local receptive fields, shared weights, and pooling. By local receptive fields, we mean that each neuron in the first(or any) hidden layer will be connected to a small region of the input(or previous layer's) neurons. Shared weights mean that we're going to use the same weights and bias for each of the local receptive field. This means that all the neurons in the hidden layer detect exactly the same feature, just at different locations in the input image. In addition to the convolutional layers just described, convolutional neural networks also contain pooling layers. Pooling layers are usually used immediately after convolutional layers. What the pooling layers do is to simplify the information in the output from the convolutional layer. Recent convolutional network architectures have 10 to 20 hidden layers and billions of connections between units. Training deep learning networks has taken weeks several years ago, but thanks to progress in GPU and algorithm enhancement, training time has reduced to several hours. Neural networks with time-varying behavior are known as recurrent neural networks or RNNs. A recurrent neural network is a class of artificial neural network where connections between units form a directed cycle. This creates an internal state of the network which allows it to exhibit dynamic temporal behavior. Unlike feedforward neural networks, RNNs can use their internal memory to process arbitrary sequences of inputs. Early RNN models turned out to be very difficult to train, harder even than deep feedforward networks. The reason is the unstable gradient problem such as vanishing gradient and exploding gradient. The gradient can get smaller and smaller as it is propagated back through layers. This makes learning in early layers extremely slow. The problem actually gets worse in RNNs, since gradients aren't just propagated backward through layers, they're propagated backward through time. If the network runs for a long time, that can make the gradient extremely unstable and hard to learn from. It has been possible to incorporate an idea known as long short-term memory units (LSTMs) into RNNs. LSTMs make it much easier to get good results when training RNNs, and many recent papers make use of LSTMs or related ideas.

• The Mathematical Education
• /
• v.27 no.1
• /
• pp.15-23
• /
• 1988

We are at the onset of a major revolution in education, a revolution unparalleled since the invention of the printing press. The computer will be the instrument of this revolution. Computers and computer application are everywhere these days. Everyone can't avoid the influence of the computer in today's world. The computer is no longer a magical, unfamiliar tool that is used only by researchers or scholars or scientists. The computer helps us do our jobs and even routine tasks more effectively and efficiently. More importantly, it gives us power never before available to solve complex problems. Mathematics instruction in secondary schools is frequently perceived to be more a amendable to the use of computers than are other areas of the school curriculum. This is based on the perception of mathematics as a subject with clearly defined objectives and outcomes that can be reliably measured by devices readily at hand or easily constructed by teachers or researchers. Because of this reason, the first large-scale computerized curriculum projects were in mathematics, and the first educational computer games were mathematics games. And now, the entire mathematics curriculum appears to be the first of the traditional school curriculum areas to be undergoing substantial trasformation because of computers. Recently, many research-Institutes of our country are going to study on computers in orders to use it in mathematics education, but the study is still start ing-step. In order to keep abreast of this trend necessity, and to enhance mathematics teaching/learning which is instructed lecture-based teaching/learning at the present time, this study aims to develop/present practical method of computer-using. This is devided into three methods. 1. Programming teaching/learning method This part is presented the following five types which can teach/learn the mathematical concepts and principle through concise program. (Type 1) Complete a program. (Type 2) Know the given program's content and predict the output. (Type 3) Write a program of the given flow-chart and solve the problem. (Type 4) Make an inference from an error message, find errors and correct them. (Type 5) Investigate complex mathematical fact through program and annotate a program. 2. Problem-solving teaching/learning method solving This part is illustrated how a computer can be used as a tool to help students solve realistic mathematical problems while simultaneously reinforcing their understanding of problem-solving processes. Here, four different problems are presented. For each problem, a four-stage problem-solving model of polya is given: Problem statement, Problem analysis, Computer program, and Looking back/Looking ahead. 3. CAI program teaching/learning method This part is developed/presented courseware of sine theorem section (Mathematics I for high school) in order to avail individualized learning or interactive learning with teacher. (Appendix I, II)