http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Adaptive Neural Network Control via Backstepping for Permanent Magnet Synchronous Motors
Sesun You,Jeonghwan Gil,Wonhee Kim 제어로봇시스템학회 2020 제어로봇시스템학회 국제학술대회 논문집 Vol.2020 No.10
In this paper, we propose an adaptive neural network control to simplify the complexity of conventional backstepping algorithm. The three-layer neural network approximator is designed to estimate the unknown continuous complex function generated by recursive backstepping process. The NN weights are updated by adaptive law based on Lyapunov stability theorem. As the weights of NN are learned according to time away, estimation error continuously decrease. The commutation scheme is applied to implement the field-oriented control without DQ transform. The performance of the proposed method was validated via simulations using MATLAB/Simulink.
ABS를 위해 휠 속 피드백 만을 이용한 Extremum-Seeking Algorithm 기반 최대 마찰력 제어
유세선(Sesun You),김원희(Wonhee Kim) 한국자동차공학회 2020 한국자동차공학회 학술대회 및 전시회 Vol.2020 No.11
In this paper, we propose a maximum friction control based on extremum-seeking algorithm using only wheel speed feedback for anti-lock braking system (ABS) in electric vehicles. In the proposed control strategy, the disturbance observer is designed to estimate the longitudinal tire-road friction. Based on the extremum-seeking algorithm, an optimal desired reference generator is designed, which provides optimal desired wheel speed in real-time to achieve the maximum tire-road friction, according to road surface conditions. A wheel speed tracking controller is developed to guarantee an upper bound of tracking errors. The performance of the proposed method is validated via simulations by using CarSim and MATLAB/Simulink for split road surface. The proposed method, which only employs a wheel speed sensor, achieves a shorter braking distance and stopping time in comparison with classical ABS.
Gwanyeon Kim,Sesun You,Jeonghwan Gil,Wonhee Kim 제어로봇시스템학회 2020 제어로봇시스템학회 국제학술대회 논문집 Vol.2020 No.10
These paper propose Electric-Power-Steering system (EPS) control method. We modified EPS dynamics from the perspective of torsion-bar torque. To robustness control with parameter uncertainty, Extended State Observer (ESO) method was applied. Additionally, The sliding mode observer which is estimated the reaction torque from the road was designed to provide the road information to the driver. The proposed method was verified by MATLAB/Simulink.
Jeonghwan Gil,Sesun You,Wonhee Kim 제어로봇시스템학회 2020 제어로봇시스템학회 국제학술대회 논문집 Vol.2020 No.10
In this paper, we propose high order sliding mode observer (HOSMO) based nonlinear gain sliding mode controller (NGSMC) for permanent magnet synchronous motors (PMSMs) under disturbance. The disturbance reduction problem is difficult to overcome in position tracking control. For achieving information about velocity and disturbances, we use HOSMO. HOSMO is guaranteed that the estimation errors are converged in the finite time. The sliding surface is designed by estimated state and measuring state. NGSMC is applied to improve position tracking performance and robustness. The proposed method is proven that tracking error is stable by Lyapunov theory and ISS property. Simulation are performed to evaluate the position tracking performance of proposed method.
슬라이딩 모드 제어기를 위한 외란관측기 설계 방법 및 분석
임재윤(Jaeyun Yim),유세선(Sesun You),김원희(Wonhee Kim) 대한전기학회 2021 대한전기학회 학술대회 논문집 Vol.2021 No.10
본 논문에서는 sliding mode controller (SMC)와 disturbance observer (DOB)의 결합에 대한 리플 증폭 문제에 대해 분석하였고, low-pass filter를 이용하여 해결 방안을 제안하였다. 제안한 DOB는 성능 검증을 위해 기존 DOB와 비교 분석을 MATLAB/Simulink를 통해 시뮬레이션 환경에서 진행하였다.
최적 노면 마찰력 기반 Anti-lock Braking System 제어
하진우(Jinwoo Ha),유세선(Sesun You),김원희(Wonhee Kim) 한국자동차공학회 2022 한국자동차공학회 부문종합 학술대회 Vol.2022 No.6
In this paper, we propose new anti-lock braking system(ABS) control technique using wheel cylinder pressure control. In the proposed control algorithm, the extended state observer estimates the longitudinal tire-road friction forces. Using these estimated value, the longitudinal tire-road friction forces are optimized by directly controlling wheel cylinder pressure based on the friction’s behavior. The performance of the proposed method is validated via hardware-in-the-loop simulation by using MATLAB/Simulink, Carsim and dSPACE vector tool for snow road. The proposed method presents a new method rather than slip ratio based control and can be further developed through various optimization algorithms.