• Journal of IKEEE
• /
• v.19 no.4
• /
• pp.541-547
• /
• 2015

In current semiconductor manufacturing, as the feature size of integrated circuit (IC) devices continuously shrinks, detecting endpoint in plasma etching process is more difficult than before. For endpoint detection, various kinds of sensors are installed in semiconductor manufacturing equipments, and sensor data are gathered with predefined sampling rate. Generally, detecting endpoint is performed using OES data of by-product. In this study, OES data of both by-product and etchant gas are used to improve reliability of endpoint detection. For the OES data pre-processing, a combination of Signal to Noise Ratio (SNR) and Principal Component Analysis (PCA),are used. Polynomial Regression and Expanded Hidden Markov model (eHMM) technique are applied to pre-processed OES data to detect endpoint.

• Journal of IKEEE
• /
• v.20 no.1
• /
• pp.9-15
• /
• 2016

The endpoint detection (EPD) is the most important technique in plasma etching process. In plasma etching process, the Optical Emission Spectroscopy (OES) is usually used to analyze plasma reaction. And Plasma Impedance Monitoring (PIM) system is used to measure the voltage, current, power, and load impedance of the supplied RF power during plasma process. In this paper, a new decision making algorithm is proposed to improve the performance of EPD in SiOx single layer plasma etching. To enhance the accuracy of the endpoint detection, both OES data and PIM data are utilized and a newly proposed decision making algorithm is applied. The proposed method successfully detected endpoint of silicon oxide plasma etching.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.22 no.2
• /
• pp.134-141
• /
• 2008

This paper presents an innovative load management technique that can effectively lower the summer peak load by adjusting the aircondition loads through smoothe coordinations between utility companies and large customers. An intelligent hierarchical load management system composed of a Central Intelligent Load Management System(CIMS) and multiple Local Intelligent Management Systems(LIMS) is also proposed to implement the reposed technique. Upon receiving a load curtailment request from the utilities, CIMS issues tokens, which can be used by each LIMS as a right to turn on the airconditioner. CIMS creates and maintains a queue for fair allocation of the tokens among the LIMS demanding tokens. By adjusting the number tokens and queue management Policies, desired load factors can be achieved conveniently. The Markov Birth and Death Process and the Balance Equations are employed in estimating various queue performances. The proposed technique is tested using a summer load data of a large apartment complex and proved to be quite effective in load management while minimizing the customer inconveniences.

• Structural Engineering and Mechanics
• /
• v.32 no.1
• /
• pp.95-126
• /
• 2009

The present contribution addresses the parallelization of advanced simulation methods for structural reliability analysis, which have recently been developed for large-scale structures with a high number of uncertain parameters. In particular, the Line Sampling method and the Subset Simulation method are considered. The proposed parallel algorithms exploit the parallelism associated with the possibility to simultaneously perform independent FE analyses. For the Line Sampling method a parallelization scheme is proposed both for the actual sampling process, and for the statistical gradient estimation method used to identify the so-called important direction of the Line Sampling scheme. Two parallelization strategies are investigated for the Subset Simulation method: the first one consists in the embarrassingly parallel advancement of distinct Markov chains; in this case the speedup is bounded by the number of chains advanced simultaneously. The second parallel Subset Simulation algorithm utilizes the concept of speculative computing. Speedup measurements in context with the FE model of a multistory building (24,000 DOFs) show the reduction of the wall-clock time to a very viable amount (<10 minutes for Line Sampling and ${\approx}$ 1 hour for Subset Simulation). The measurements, conducted on clusters of multi-core nodes, also indicate a strong sensitivity of the parallel performance to the load level of the nodes, in terms of the number of simultaneously used cores. This performance degradation is related to memory bottlenecks during the modal analysis required during each FE analysis.

• The Journal of the Korea Contents Association
• /
• v.5 no.1
• /
• pp.147-155
• /
• 2005

In this paper, communication channels for the transmission of multimedia packets are modeled and evaluated. The multimedia packet traffic characterized by on-off and MMPP process for voice and data, respectively, dynamic channel allocation, queueing of data packets due to unavailability of channels and dropping of queued data packets over timeout, and guard channel for voice packets are modeled. The performance indices adopted in the evaluation of SRN model includes blocking and dropping probabilities. The SRN uses rewards concepts instead of the complicate numerical analysis required for the Markov chain. It is shown that our SRN modeling techniques provide an easier way to carry out performance analysis.

• Journal of the Korea Computer Industry Society
• /
• v.2 no.11
• /
• pp.1395-1406
• /
• 2001

In order to cover a widespread service range, terrestrial/satellite-mixed network is being combined with terrestrial ATM network. This dissertation analyzes and investigates several previously existent CDV compensation methods in order to compensate CDV arising from interfacing satellite TDMA and ATM. Specifically to supplement the problems of timestamp and cell number counting methods, new Variable Timestamp method for CDV compensation is proposed. To evaluate the proposed method, MMPP(Markov Modulated Poisson Process), which can express VBR service very well, is selected as a cell input traffic model of terrestrial transmitting earth station. After several simulation, it is also confirmed that CDV compensation capability for VBR services is very superior to the cell number counting method. In this case, as the timestamp number Nts increases, CDV compensation capability increases, and the CDV distribution length is reduced.

• The Korean Journal of Applied Statistics
• /
• v.27 no.6
• /
• pp.959-973
• /
• 2014

Asset pricing models that account for asymmetric volatility in asset prices have been recently proposed. This article presents an efficient Bayesian method to analyze asset-pricing models. The method is developed by devising a partially collapsed Gibbs sampler that capitalizes on the functional incompatibility of conditional distributions without complicating the updates of model components. The proposed method is illustrated using simulated data and applied to daily S&P 500 data observed from September 1980 to August 2014.

• Journal of Korea Multimedia Society
• /
• v.5 no.5
• /
• pp.523-529
• /
• 2002

The present study describes neural networks for stereoscopic vision, which are applied to identifying human speech. In speech recognition based on stereoscopic vision neural networks (SVNN), the similarities are first obtained by comparing input vocal signals with standard models. They are then given to a dynamic process in which both competitive and cooperative processes are conducted among neighboring similarities. Through the dynamic processes, only one winner neuron is finally detected. In a comparative study, the two-layered SVNN was 7.7% higher in recognition accuracies than the hidden Markov model (HMM). From the evaluation results, it was noticed that SVNN outperformed the existing HMM recognizer.

• Journal of the Korean Society of Industry Convergence
• /
• v.18 no.4
• /
• pp.250-258
• /
• 2015

In this research, we proposed a new approach to implement the real-time motion control of biped robot based on voice command for unmanned FA. Voice is one of convenient methods to communicate between human and robots. To command a lot of robot task by voice, voice of the same number have to be able to be recognition voice is, the higher the time of recognition is. In this paper, a practical voice recognition system which can recognition a lot of task commands is proposed. The proposed system consists of a general purpose microprocessor and a useful voice recognition processor which can recognize a limited number of voice patterns. Given biped robots, each robot task is, classified and organized such that the number of robot tasks under each directory is net more than the maximum recognition number of the voice recognition processor so that robot tasks under each directory can be distinguished by the voice recognition command. By simulation and experiment, it was illustrated the reliability of voice recognition rates for application of the manufacturing process.

• The Journal of Korea Robotics Society
• /
• v.17 no.1
• /
• pp.76-85
• /
• 2022

One of the most fundamental challenges when designing controllers for dynamic systems is the adjustment of controller parameters. Usually the system model is used to get the initial controller, but eventually the controller parameters must be manually adjusted in the real system to achieve the best performance. To avoid this manual tuning step, data-driven methods such as machine learning were used. Recently, reinforcement learning became one alternative of this problem to be considered as an agent learns policies in large state space with trial-and-error Markov Decision Process (MDP) which is widely used in the field of robotics. However, on initial training step, as an agent tries to explore to the new state space with random action and acts directly on the controller parameters in real systems, MDP can lead the system safety-critical system failures. Therefore, the issue of 'safe exploration' became important. In this paper we meet 'safe exploration' condition with Control Barrier Function (CBF) which converts direct constraints on the state space to the implicit constraint of the control inputs. Given an initial low-performance controller, it automatically optimizes the parameters of the control law while ensuring safety by the CBF so that the agent can learn how to predict and control unknown and often stochastic environments. Simulation results on a quadrotor UAV indicate that the proposed method can safely optimize controller parameters quickly and automatically.