• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2004.04a
• /
• pp.125-128
• /
• 2004

Q-learning은 강화학습의 한 방법으로서, 여러 분야에 널리 응용되고 있는 기법이다. 최근에는 Linear Quadratic Regulation (이하 LQR) 문제에 성공적으로 적용된 바 있다. 특히 시스템 모델의 파라미터에 대한 구체적인 정보가 없는 상태에서 적절한 입력과 출력만을 가지고, 학습을 통해 문제를 해결할 수 있어서 상황에 따라서 매우 실용적인 대안이 될 수 있다. 이에 따라 본 논문에서는 이러한 일반적인 LQR Q-learning(이하 LQRQL) 학습방법에 퍼지 모델을 이용하여 제어기를 설계하는 방법을 고려하고, 일반적인 LQROL 기법과 본 논문에서 제시한 방법의 결과를 비교하여 응용 가능성을 살펴보았다.

• Journal of Mechanical Science and Technology
• /
• v.17 no.3
• /
• pp.329-335
• /
• 2003

EALQR (Linear Quadratic Regulate. design with Eigenstructure Assignment capability) has the capability of exact assignment of eigenstructure with the guaranteed margins of the LQR for MIMO (Multi-Input Multi-Output) systems. However, EALQR undergoes a restriction on the state-weighting matrix Q in LQR to be indefinite with respect to the region of allowable closedloop eigenvalues. The definiteness of the weighting matrix is closely related to the robustness property of a given system. In this paper, we derive a relation between the indefinite weighting matrix Q and the robustness property for EALQR. The modified frequency domain inequality, that could be guaranteed by EQLQR with an indefinite weighting matrix, is presented.

• Proceedings of the Korean Society of Propulsion Engineers Conference
• /
• 1998.10a
• /
• pp.37-37
• /
• 1998

중형항공기용 터보펜 엔진의 성능모사와 LQR 제어기 설계에 대한 연구를 수행하였다. 동적 성능모사를 설계점으로 선정한 지상최대이륙조건과 탈설계점으로 선정한 최대상승조건과 순항조건에 대하여 Step 증가, Ramp 증가, Ramp 감소, Step증가 후 Ramp 감소의 4가지 연료공급에 조건에 대하여 수행되었다. 성능모사 결과 모든 비행조건에서 연료를 Step 증가시킬 경우 고압터보빈의 입구온도가 제한온도인 3105$^{\circ}$R을 초과함을 확인하였고, 최대 상승조건에서 연료를 Step 증가시킬 경우가 4.5초 이내에 Ramp 증가시킬 경우 고압압축기에 서지가 발생함을 확인하였다. 따라서 고압터어빈의 오버슈트와 고압압축기의 서지를 동시에 제어할 수 있는 다변수 제어기의 설계가 필요함을 확인하였다.

• 한국신재생에너지학회:학술대회논문집
• /
• 2011.11a
• /
• pp.40.1-40.1
• /
• 2011

This paper presents a pitch controller which can hold output power constant at the rated value. Although wind turbine contains complicated nonlinearities, its behaviour within a certain operating range of a point can be approximated by that of a linear model. By doing so, we can apply rather simple and systematic linear control techniques such as PID and LQR(Linear Quadratic Regulator) to design a linear pitch controller. Because these linear controllers are valid only in a sufficiently small range around an operating point, linearized wind turbine model under the condition of varying wind speed needs a linear pitch controller can achieve the aims of tracking the rated rotor rotational speed. We propose an improved linear pitch controller taking each merit of LQR and PI controller under the condition of varying operating points in this paper.

• Journal of Institute of Control, Robotics and Systems
• /
• v.18 no.7
• /
• pp.617-621
• /
• 2012

This paper studies asymptotic consensus problem for linear multi-agent systems. We propose a distributed state feedback control algorithm for solving the problem under fixed and undirected network communication. In contrast with the conventional algorithms that use global information (e.g., graph connectivity), the proposed algorithm only uses local information from neighbors. The principle for achieving asymptotic consensus is that, for each agent, a distributed update law gradually increases the coupling gain of LQR-type feedback and thus, the overall stability of the multi-agent system is recovered by the gain margin of LQR.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.12 no.5
• /
• pp.311-318
• /
• 2017

A reduced order observer is developed for depth control of a hybrid underwater glider which combines the good aspects of a conventional autonomous underwater vehicle and a underwater glider. State variables include the center of gravity of the robot and the weight of the buoyancy bag, which can not be directly measured. By using the mathematical model and available information such as directional velocities, accelerations, and attitudes, we developed a Luenberger's reduced order observer to estimate the center of gravity and the buoyancy weight. By simulations using Matlab/Simulink, the efficiency of the proposed observer is shown, where a LQR controller using full state variables is adopted as a depth controller.

• Journal of Mechanical Science and Technology
• /
• v.15 no.6
• /
• pp.776-787
• /
• 2001

The turbofan engine performance analysis for a medium scale commercial aircraft was carried out and the LQR control scheme for performance optimization was studied. By using scaled component maps from well-known CF6 engine characteristics, the steady-state performance analysis result was compared with BR715-56 engine performance data. The transient performance analysis was performed with four fuel schedules. The linear simulation was done at the maximum take-off condition. The real time linear simulation was performed by interpolation of the system matrices, which used the least square method as the function of LPC rotational speed. By using linear system matrices of design point, the LQR controller which used control variables for the fuel flow and the LPC bleed air was designed.

• Journal of the Korean Society for Precision Engineering
• /
• v.20 no.5
• /
• pp.99-107
• /
• 2003

In a winding process, important control specifications include regulation of web tension and velocity. In this research, an adaptive controller has been developed for controlling web tension and velocity in winding processes. For the controller design, the linear quadratic regulator theory has been adopted and a gain-scheduling scheme has been incorporated. A prototype winding system has been constructed, and the controller has been implemented in a real-time PC-based environment. The performance of the closed loop system has been evaluated via simulation and experiments, and it was observed that both the web tension and velocity could be regulated within a small tolerance.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.15 no.5
• /
• pp.352-360
• /
• 2010

This paper proposes a precise position control of an automatic door using a disturbance observer. Although the conventional PID controller is usually used for an automatic door system, the demand of the robust controller considering the parameter deviations and disturbances is increasing due to the various size and weight of a door. The linear model for an automatic door system is presented. Based on this model, the LQR controller using a state feedback controller and an observer are suggested. A disturbance observer to compensate the undesirable factors is also proposed. Simulation and Experimental results are presented to illustrate the feasibility of the proposed control strategy.

• Journal of Institute of Control, Robotics and Systems
• /
• v.19 no.1
• /
• pp.1-9
• /
• 2013

This paper presents a new control algorithm for dynamic control of a unicycle robot. The unicycle robot motion consists of a pitch that is controlled by an in-wheel motor and a roll that is controlled by a reaction wheel pendulum. The unicycle robot doesn't have any actuator for a yaw axis control, which makes the derivation of the dynamics relatively simple. The Euler-Lagrange equation is applied to derive the dynamic equations of the unicycle robot to implement the dynamic speed control of the unicycle robot. To achieve the real time speed control of the unicycle robot, the sliding mode control and LQ regulator are utilized to guarantee the stability while maintaining the desired speed tracking performance. In the roll controller, the sigmoid-function based sliding mode controller has been adopted to minimize the chattering by the switching function. The LQR controller has been implemented for the pitch control to drive the unicycle robot to follow the desired velocity trajectory in real time using the state variables of pitch angle, angular velocity, angle and angular velocity of the wheel. The control performance of the two control systems form a single dynamic model has been demonstrated by the real experiments.