Cheol Hoon Park

Korea Institute of Machinery and Materials, Sŏul, Seoul, South Korea

Are you Cheol Hoon Park?

Claim your profile

Publications (63)49.79 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: The development of long-duration mobile energy sources suitable for mobile robots or small-sized unmanned vehicles is an area of active research. The micro gas turbine generator (MTG) is a promising solution for this purpose because it has both a high energy density and a high power density. A 500W class MTG is under development. The design of the MTG can be divided into two main parts. One part consists of the motor/generator and compressor, whereas the other part consists of the combustor, turbine, and recuperator. The 500W class MTG is designed to operate at an extremely high speed of 400,000 rpm, with a turbine temperature of over 700 °C, to improve the efficiency. Because the magnetism of the NdFeB permanent magnet for the motor/generator would degrade if it were exposed to temperatures over 150-200 °C, thermal insulation between the two parts is needed to reduce the heat transfer from the combustor and turbine. The motor/generator is also cooled by the inlet air flow driven by the compressor. This study presents the results of an experiment designed to verify whether the thermal insulation and air flow are effective enough to keep the temperature of the motor/generator part below 100 °C. From the rotating test, performed with the aid of a high-temperature test rig, it was confirmed that a temperature of less than 50 °C could be maintained in the motor/generator part, under the condition of 1.0 bar compressed air supply.
    2014 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM); 07/2014
  • [Show abstract] [Hide abstract]
    ABSTRACT: This Note proposes a novel radial-thrust combo metal mesh foil bearing (MMFB). Although MMFBs have advantages such as higher stiffness and damping over conventional air foil bearings, studies related to MMFBs have been limited to radial MMFBs. The novel combo MMFB is composed of a radial top foil, thrust top foils, and a ring-shaped metal mesh damper-fabricated by compressing a copper wire mesh-with metal mesh thrust pads for the thrust bearing at both side faces. In this study, the combo MMFB was fabricated in half-split type to support the rotor for a micro gas turbine generator. The manufacture and assembly process for the half-split-type combo MMFB is presented. In addition, to verify the proposed combo MMFB, motoring test results up to 250 000 rpm and axial displacements as a function of rotational speed are presented.
    The Review of scientific instruments 10/2013; 84(10):106102. · 1.52 Impact Factor
  • Cheol Hoon Park, Soohyun Kim, Kyung-Soo Kim
    [Show abstract] [Hide abstract]
    ABSTRACT: Centrifuges are devices that separate particles of different densities and sizes through the application of a centrifugal force. If a centrifuge could be operated under atmospheric conditions, all vacuum-related components such as the vacuum chamber, vacuum pump, diffusion pump, and sealing could be removed from a conventional centrifuge system. The design and manufacturing procedure for centrifuges could then be greatly simplified to facilitate the production of lightweight centrifuge systems of smaller volume. Furthermore, the maintenance costs incurred owing to wear and tear due to conventional ball bearings would be eliminated. In this study, we describe a novel vacuum chamber-free centrifuge supported by magnetic bearings. We demonstrate the feasibility of the vacuum chamber-free centrifuge by presenting experimental results that verify its high-speed support capability and motoring power capacity.
    The Review of scientific instruments 09/2013; 84(9):095106. · 1.52 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Hybrid-type magnetic bearings using both permanent magnets and electromagnets have been used for rotating machinery. In the case of conventional thrust hybrid magnetic bearings supporting axial loads, radially magnetized permanent ring magnets, which have several demerits such as difficult magnetization and assembly, have been used to generate bias flux. In this study, a novel thrust hybrid magnetic bearing using an axially magnetized permanent ring magnet is presented. Because it is easy to magnetize a ring magnet in the axial direction, the segmentation of the ring magnet for magnetization is not required and the assembly process can be simplified. For verifying the performance of the proposed method, a test rig that consists of a proposed thrust magnetic bearing and variable loads is constructed. This paper presents the detailed design procedures and the obtained experimental results. The results show that the developed thrust magnetic bearing has the potential to replace conventional thrust magnetic bearings.
    Journal of Magnetics 01/2013; 18(3). · 0.33 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Satellites provide a lot of information and essay roles in the areas of defense and space observations. The precise distances to the satellites are measured by emitting and retro-reflecting a laser. For such surveys, satellite laser ranging (SLR) systems have been developed in different forms and for different areas. The structural integrity of the tracking mount is essential for it to be able to track a high-speed satellite precisely, overcoming the various external and internal disturbances and operating conditions. In this study, the analysis of a tracking mount was performed for weight, wind loads, and inertia loads in order to verify its soundness. The results of the comparison between aluminum and steel were analyzed in order to select the optimal material for the fork and main housing part. In addition, the natural frequency and mode shape were predicted. Optimal material selection and structural integrity will also be verified using static analysis.
    Journal of manufacturing engineering & technology. 01/2013; 22(5).
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this study, we have proposed and implemented a design for the tracking mount and controller of the ARGO-M (Accurate Ranging system for Geodetic Observation – Mobile) which is a mobile satellite laser ranging (SLR) system developed by the Korea Astronomy and Space Science Institute (KASI) and Korea Institute of Machinery and Materials (KIMM). The tracking mount comprises a few core components such as bearings, driving motors and encoders. These components were selected as per the technical specifications for the tracking mount of the ARGO-M. A three-dimensional model of the tracking mount was designed. The frequency analysis of the model predicted that the first natural frequency of the designed tracking mount was high enough. The tracking controller is simulated using MATLAB/xPC Target to achieve the required pointing and tracking accuracy. In order to evaluate the system repeatability and tracking accuracy of the tracking mount, a prototype of the ARGO-M was fabricated, and repeatability tests were carried out using a laser interferometer. Tracking tests were conducted using the trajectories of low earth orbit (LEO) and high earth orbit (HEO) satellites. Based on the test results, it was confirmed that the prototype of the tracking mount and controller of the ARGO-M could achieve the required repeatability along with a tracking accuracy of less than 1arcsec.
    Advances in Space Research 01/2012; · 1.18 Impact Factor
  • Source
    Yuxi Hou, Iickho Song, Hwang-Ki Min, Cheol Hoon Park
    [Show abstract] [Hide abstract]
    ABSTRACT: Owing to the singularity of the within-class scatter, linear discriminant analysis (LDA) becomes ill-posed for small sample size (SSS) problems. Null-space-based LDA (NLDA), which is an extension of LDA, provides good discriminant performances for SSS problems. Yet, as the original scheme for the feature extractor (FE) of NLDA suffers from a complexity burden, a few modified schemes have since been proposed for complexity reduction. In this brief, by transforming the problem of finding the FE of NLDA into a linear equation problem, a novel scheme is derived, offering a further reduction of the complexity.
    IEEE transactions on neural networks and learning systems 01/2012; 23(6):1003-1009. · 4.37 Impact Factor
  • Source
    Jong-Seok Lee, Cheol Hoon Park
    [Show abstract] [Hide abstract]
    ABSTRACT: We propose a novel stochastic optimization algorithm, hybrid simulated annealing (SA), to train hidden Markov models (HMMs) for visual speech recognition. In our algorithm, SA is combined with a local optimization operator that substitutes a better solution for the current one to improve the convergence speed and the quality of solutions. We mathematically prove that the sequence of the objective values converges in probability to the global optimum in the algorithm. The algorithm is applied to train HMMs that are used as visual speech recognizers. While the popular training method of HMMs, the expectation-maximization algorithm, achieves only local optima in the parameter space, the proposed method can perform global optimization of the parameters of HMMs and thereby obtain solutions yielding improved recognition performance. The superiority of the proposed algorithm to the conventional ones is demonstrated via isolated word recognition experiments.
    IEEE Transactions on Systems Man and Cybernetics Part B (Cybernetics) 09/2010; · 3.24 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: A class of recurrent neural networks is proposed and proven to be capable of identifying any discrete-time dynamical system. The application of the proposed network is addressed in the encoding, identification, and extraction of finite state automata (FSAs). Simulation results show that the identification of FSAs using the proposed network, trained by the hybrid greedy simulated annealing with a modified cost function in the training stage, generally exhibits better performance than the conventional identification procedures.
    IEEE Transactions on Neural Networks 09/2010; 21(9):1408-21. · 2.95 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Various passive and/or active antirolling devices have been used for suppressing the rolling motion of ships in the ocean. In this study, a maglev-type active mass driver (AMD) is developed for controlling the rolling motion of a shiplike structure. No friction is generated during the motion of this maglev-type AMD, as the moving mass is floated by the magnetic levitation force and displaced by the propulsion force generated by the linear motor. For verifying the feasibility of the proposed method, a small AMD having a moving mass of approximately 4.0 kg is constructed and used in a small-scale model of a catamaran. This paper presents the detailed design procedures and obtained experimental results. Our results show that the developed maglev-type AMD has the potential for use in controlling the rolling motion of ships and other oceanographic vessels.
    The Review of scientific instruments 05/2010; 81(5):056102. · 1.52 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: A 5 kWh class FESS (flywheel energy storage system) with the operating speed range of 9,000~15,000 rpm has been developed. The system consists of a composite flywheel rotor, active magnetic bearings, a motor/generator and its controller. Because active magnetic bearings(AMB) to support the rotor vibration are open-loop unstable, they needs a feedback controller for stable operation. Digital controller and notch filters have been designed and implemented on a MATLAB/Simulink and xPC target with 10 kHz sample rate. MATLAB/xPC target which is PC based real-time controller is useful to get FRF (frequency response function) of AMB and implement the notch filters and the control algorithms. This paper introduces the FESS mounted on AMBs being controlled by xPC system and the experimental results up to 15,000 rpm.
    Computer Science and Information Engineering, 2009 WRI World Congress on; 05/2009
  • Hajoon Lee, Cheol Hoon Park
    [Show abstract] [Hide abstract]
    ABSTRACT: We deal with LTI nonminimum phase (NMP) systems which are difficult to control with conventional methods because of their inherent characteristics of undershoot. In such systems, reducing the undesirable undershoot phenomenon makes the response time of the systems much longer. Moreover, it is impossible to control the magnitude of undershoot in a direct way and to predict the response time. In this paper, we propose a novel two sliding mode control scheme which is capable of stably determining the magnitude of undershoot and thus the response time of NMP systems a priori. To do this, we introduce two sliding lines which are in charge of control in turn. One is used to stabilize the system and achieve asymptotic regulation eventually like the conventional sliding mode methods and the other to stably control the magnitude of undershoot from the beginning of control until the state meets the first sliding line. This control scheme will be proved to have an asymptotic regulation property. The computer simulation shows that the proposed control scheme is very effective and suitable for controlling the NMP systems compared with the conventional ones.
    IEICE Transactions. 01/2009; 92-A:862-870.
  • [Show abstract] [Hide abstract]
    ABSTRACT: The aortic pressure (AoP) and pulmonary artery pressure (PAP) are important factors for the long-term survival of the patient receiving an implant of a total artificial heart (TAH). A new afterload regulation method using neural controllers has been developed for a moving actuator type TAH. Without any transducer, the proposed neural controller, based on the predetermined peak level of the motor current, regulates AoP by adjusting the actuator velocity. At the same time, a fuzzy controller based on expert knowledge prevents the PAP from being abnormally high. The proposed controller not only regulates the pump output on the maintenance of the normal AoP and PAP, but it also successfully controls the suction in the right heart. In vitro tests show that the proposed neural and fuzzy controllers effectively regulate AoP and PAP as well as prevent atrial suction.
    Artificial Organs 11/2008; 20(11):1220 - 1226. · 1.96 Impact Factor
  • Source
    Jong-Seok Lee, Cheol Hoon Park
    [Show abstract] [Hide abstract]
    ABSTRACT: This chapter addressed the problem of information fusion for AVSR. We introduced the bimodal nature of speech production and perception by humans and defined the goal of audio-visual integration. We reviewed two existing approaches for implementing audiovisual fusion in AVSR systems and explained the preference of decision fusion to feature fusion for constructing noise-robust AVSR systems. For implementing a noise-robust AVSR system, different definitions of the reliability of a modality were discussed and compared. A neural network-based fusion method was described for effectively utilizing the reliability measures of the two modalities and producing noise-robust recognition performance over various noise conditions. It has been shown that we could successfully obtain the synergy of the two modalities. The audio-visual information fusion method shown in this chapter mainly aims at obtaining robust speech recognition performance, which may lack modelling of complicated humans' audio-visual speech perception processes. If we consider that the humans' speech
    Speech Recognition, 11/2008; , ISBN: 978-953-7619-29-9
  • Jong-Seok Lee, Cheol Hoon Park
    [Show abstract] [Hide abstract]
    ABSTRACT: Audio-visual speech recognition (AVSR) using acoustic and visual signals of speech has received attention because of its robustness in noisy environments. In this paper, we present a late integration scheme-based AVSR system whose robustness under various noise conditions is improved by enhancing the performance of the three parts composing the system. First, we improve the performance of the visual subsystem by using the stochastic optimization method for the hidden Markov models as the speech recognizer. Second, we propose a new method of considering dynamic characteristics of speech for improved robustness of the acoustic subsystem. Third, the acoustic and the visual subsystems are effectively integrated to produce final robust recognition results by using neural networks. We demonstrate the performance of the proposed methods via speaker-independent isolated word recognition experiments. The results show that the proposed system improves robustness over the conventional system under various noise conditions without a priori knowledge about the noise contained in the speech.
    IEEE Transactions on Multimedia 09/2008; · 1.75 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we present a real-time system to detect abnormal events on gas pipes, based on the signals which are observed through the audio sensors attached on them. First, features are extracted from this signal so that they are robust to noise and invariant to the distance between a sensor and a spot at which an abnormal event like an attack on the gas pipes occurs. Then, a classifier is constructed to detect abnormal events using neural networks. It is a combination of two neural network models, a Gaussian mixture model and a multi-layer perceptron, for the reduction of miss and false alarms. The former works for miss alarm prevention and the latter for false alarm prevention. The experimental result with real data from the actual gas system shows that the propose system is effective in detecting the dangerous events in real-time having an accuracy of 92.9%.
    Industrial Electronics Society, 2007. IECON 2007. 33rd Annual Conference of the IEEE; 12/2007
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper presents a single-objective and a multiobjective stochastic optimization algorithms for global training of neural networks based on simulated annealing. The algorithms overcome the limitation of local optimization by the conventional gradient-based training methods and perform global optimization of the weights of the neural networks. Especially, the multiobjective training algorithm is designed to enhance generalization capability of the trained networks by minimizing the training error and the dynamic range of the network weights simultaneously. For fast convergence and good solution quality of the algorithms, we suggest the hybrid simulated annealing algorithm with the gradient-based local optimization method. Experimental results show that the per- formance of the trained networks by the proposed methods is better than that by the gradient-based local training algorithm and, moreover, the generalization capability of the networks is significantly improved by preventing overfitting phenomena.
    Proceedings of the IEEE Congress on Evolutionary Computation, CEC 2007, 25-28 September 2007, Singapore; 01/2007
  • Jong-Seok Lee, Cheol Hoon Park
    [Show abstract] [Hide abstract]
    ABSTRACT: ABSTRACT The use of visual ,information of speech ,has been shown ,to be effective for compensating ,for ,performance ,degradation ,of acoustic speech recognition in noisy environments. However, visual noise is usually ,ignored in most ,of audio-visual ,speech recognition systems, while it can be included in visual speech signals during acquisition or transmission ,of the ,signals. In this paper, we present a new temporal filtering technique for extraction of noise-robust visual features. In the proposed method, acarefully,designed ,band-pass filter is applied ,to the ,temporal pixel value sequences ,of lip ,region images ,in order ,to remove unwanted temporal variations due to visual noise, illumination conditions or speakers’ ,appearances. We demonstrate ,that the method,can ,improve ,not ,only ,visual ,speech ,recognition performance,for clean and noisy images ,but also audio-visual speech recognition performance,in both,acoustically and visually noisy conditions. Categories and Subject Descriptors I.4.7 [Image Processing and ,Computer ,Vision]: Feature measurement,– ,feature ,representation; I.2.7 [Artificial Intelligence]: Natural Language Processing – speech recognition and synthesis General Terms
    Proceedings of the 9th International Conference on Multimodal Interfaces, ICMI 2007, Nagoya, Aichi, Japan, November 12-15, 2007; 01/2007
  • Jong-Seok Lee, Cheol Hoon Park
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper presents a novel training algorithm of hidden Markov models (HMMs) for visual speech recognition based on a modified simulated annealing (SA) algorithm, hybrid simulated annealing, where SA is combined with a local optimization technique to improve the convergence speed and the solution quality. While the popular training method of HMMs, the expectation-maximization (EM) algorithm, only achieves local optima in the parameter space, the proposed algorithm performs global search and thus obtains solutions giving improved recognition performance. The effectiveness of the proposed method is demonstrated via isolated word recognition experiments.
    Systems, Man and Cybernetics, 2006. SMC '06. IEEE International Conference on; 11/2006
  • Dingjun Chen, Cheol Hoon Park
    [Show abstract] [Hide abstract]
    ABSTRACT: A very effective way in identifying the global optimum solution to a difficult function optimization problem is to effectively combine genetic algorithms (GA) with simulated annealing (SA) and build hybrid genetic algorithms and simulated annealing (HGASA). We implement a sequential HGASA and measure its performance; the performance survey shows that it is a very good method as compared with some sequential optimization algorithms that offer low efficiency and limited reliability. However, the sequential HGASA generally needs a long run time cost. Thus we implement a parallel HGASA using the message passing interface (MPI) on high performance computers and perform many tests using a set of frequently used function optimization problems. The detailed performance analysis of this parallel approach is done as well in terms of program execution time, relative speed up and efficiency. The experimental results show that this parallel optimization method is very suitable for tackling some complicated function optimization problems.
    Pacific Journal of Optimization 01/2006; 2(1). · 0.70 Impact Factor

Publication Stats

236 Citations
49.79 Total Impact Points

Institutions

  • 2009–2012
    • Korea Institute of Machinery and Materials
      Sŏul, Seoul, South Korea
  • 1993–2010
    • Korea Advanced Institute of Science and Technology
      • Department of Electrical Engineering
      Seoul, Seoul, South Korea
  • 2005
    • Korean Institute of Ocean Science and Technology
      Anzan, Gyeonggi Province, South Korea
  • 2002
    • Harvard University
      • School of Engineering and Applied Sciences
      Cambridge, MA, United States