http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
柳祥桓,洪錫敎,高永吉 亞洲大學校 1987 論文集 Vol.10 No.-
This paper deals with the decision of 7 primitive motions of the arm by EMG pattern recognition using IMB PC. The EMG signals are detected from silver/silver Chloride surface electrodes, and IBM PC calculates the likelyhood probability and the decision function on the feature space of integral absolute value. Multiclass decision rule and learning algorithm are introduced for high decision rate. All procedures are programmed in assembly language for the real time implementation and the co-processor 8087(numeric data processor) is used for the exponential function and the probability calculation. On our experimental results from the experted simulator, the decision rate of more than 90 percent can be obtained.
白祥基,高永吉,洪錫敎 亞州大學校 1990 論文集 Vol.13 No.-
Two methods for reducing the position error for robot are discussed in this paper. The first one which is called the compensation method compensates the parameters of the kinematic equation using the error between the real and caculated end positions. The other one,the teaching method,modifies the joint variables using the informations of the teaching points. The application to the SCARA type robot shows that the position error can be redced from 2.7mm to 0.7mm.
쿨롱 마찰력 추정과 보상을 통한 역진자 시스템의 제어 성능 개선
朴德基(Duckgee Park),左東京(Dongkyoung Chwa),洪錫敎(Suk-Kyo Hong) 대한전기학회 2006 전기학회논문지 D Vol.55 No.11
When the nonlinearities, such as friction and backlash, are not considered in the controller design, undesirable oscillations can occur in the steady-state response of a control system. This paper deals with a method to reduce oscillations that often appear in the steady-state response of a pendulum system, which is controlled by a state feedback controller based on the linearized system model. With an assumption that the oscillations shown in the steady-state are caused by the Coulomb friction, we improve the performance of stabilization and tracking by estimating and compensating for the Coulomb friction in the pendulum system. Experimental results show that the control performance can be improved sufficiently by the proposed method, when it is applied to an inverted cart pendulum which is a multi-variable unstable system. Furthermore, we could see that the Coulomb friction model used in the estimation of the friction is valid in applying the suggested method.