• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2004.10a
• /
• pp.211-216
• /
• 2004

In this thesis, we present the Takagi-Sugeno-Kang (TSK) fuzzy model based adaptive controller and adaptive identification for a general class of uncertain nonlinear dynamic systems. We use an estimated model for the unknown plant model and use this model for designing the controller. The hybrid adaptive control combined direct and indirect adaptive control based on TSK fuzzy model is constructed. The direct adaptive law can be showed by ignoring the identification errors and fails to achieve parameter convergence. Thus, we propose an TSK fuzzy model based hybrid adaptive (HA) law combined of the tracking error and the model ins error to adjust the parameters. Using a Lyapunov synthesis approach, the proposed hybrid adaptive control is proved. The hybrid adaptive law (HA) is better than the direct adaptive (DA) method without identifying the model ins error in terms of faster and improved tracking and parameter convergence. In order to show the applicability of the proposed method, it is applied to the inverted pendulum system and the performance is verified by some simulation results.

• Journal of Mechanical Science and Technology
• /
• v.20 no.5
• /
• pp.602-611
• /
• 2006

This paper proposes a locomotion pattern and a control method for biped robots with curved soles. First, since the contact point of a supporting leg may arbitrarily move back and forth on the ground, we derived the desired trajectory from a model called the Moving. Inverted Pendulum Model (MIPM) where the Zero Moment Point (ZMP) exists at the supporting point and can be moved intentionally. Secondly, a biped robot with curved soles is an under-actuated system since the supporting point contacting with a point on the ground has no actuator during the single supporting phase. Therefore, this paper proposes a computed-torque control for this under-actuated system using decoupled dynamic equations. A series of computer simulations with a 7-DOF biped robot with curved soles shows that the proposed walking pattern and control method are effective and allow the biped robot to walk fast and stably, and move more like human beings. Also, it is shown that the curved sole shape has superior energy consumption compared to flat soles, and greater efficiency in ascending and descending the stairs.

• The Journal of Korea Robotics Society
• /
• v.11 no.3
• /
• pp.183-192
• /
• 2016

Generating motion of center of mass for biped robots is a challenging issue since biped robots can easily lose balance due to limited contact area between foot and ground. In this paper, we propose force control method to generate high-speed motion of the center of mass for horizontal direction without losing balancing condition. Contact consistent multi-body dynamics of the robot is used to calculate force for horizontal direction of the center of mass considering balance. The calculated force is applied for acceleration or deceleration of the center of mass to generate high speed motion. The linear inverted pendulum model is used to estimate motion of the center of mass and the estimated motion is used to select either maximum or minimum force to stop at goal position. The proposed method is verified by experiments using 12-DOF torque controlled human sized legged robot.

• Advances in robotics research
• /
• v.2 no.2
• /
• pp.151-159
• /
• 2018

The evolution of bipedal robots was the foundation stone for development of Humanoid robots. The highly complex and non-linear dynamic of human walking made it very difficult for researchers to simulate the gait patterns under different conditions. Simple controllers were developed initially using basic mechanics like Linear Inverted Pendulum (LIP) model and later on advanced into complex control systems with dynamic stability with the help of high accuracy feedback systems and efficient real-time optimization algorithms. This paper illustrates a number of significant mathematical models and controllers developed so far in the field of bipeds and humanoids. The key facts and ideas are extracted and categorized in order to describe it in a comprehensible structure.

• Journal of Multimedia Information System
• /
• v.2 no.1
• /
• pp.179-186
• /
• 2015

This paper describes the application of a simulated annealing to a random signal-based learning. The simulated annealing is used to generate the reinforcement signal which is used in the random signal-based learning. Random signal-based learning is similar to the reinforcement learning of neural network. It is poor at hill-climbing, whereas simulated annealing has an ability of probabilistic hill-climbing. Therefore, hybridizing a random signal-based learning with the simulated annealing can produce better performance than before. The validity of the proposed algorithm is confirmed by applying it to two different examples. One is finding the minimum of the nonlinear function. And the other is the optimization of fuzzy control rules using inverted pendulum.

• Proceedings of the KIEE Conference
• /
• 1999.11c
• /
• pp.579-581
• /
• 1999

본 논문에서는 컴퓨터의 정보처리 능력과 시각기능인 CCD 카메라의 영상처리 능력을 결합시켜 극한 상황에서도 실제 시스템을 효과적으로 제어할 수 있는 실시간 비전시각 제어시스템을 제안하고, 이를 대표적인 벤치마컴 시스템인 도립진자 시스템에 적용하여 실증하였다. 우선, 전용화 된 하드웨어를 사용하지 않고. 컴퓨터를 직접 사용하므로 영상처리 중에 발생하는 많은 데이터에서 필요한 정보를 신속하게 획득하고 처리할 수 있는 새로운 알고리즘을 제안하고 이를 시뮬레이션을 통하여 검증하였다. 또한, 실제 비전 제어시스템을 제작하고, 제안된 알고리즘을 비선형 도립진자의 제어에 적용하여 퍼지 제어기를 설계하므로 컴퓨터를 이용한 실시간 비전 시각 영상처리 제어의 가능성과 우수성을 입증하였다.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.2538-2541
• /
• 2005

This paper deals one of the methods of system identification, especially on-line system identification in time-domain. The algorithm in this study needs all states of the system as well input to it for system identification. In this reason, Kalman filter is used for state estimation. But in order to implement a state estimator, the fact that a system model must be known is logical contradiction. To overcome this, state estimation and system parameter estimation are performed simultaneously in one sample. And the result of the system parameter estimation is used as basis to state estimation in next sample. On-line system identification comes, in every sample by performing both processes of state estimation and parameter estimation that are related mutually and recursively. This paper demonstrates the validity of proposed algorithm through an example of an unstable inverted pendulum system. This algorithm can be useful for on-line system identification of a system that has fewer number of measurable output than system order or number of states.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.8 no.5
• /
• pp.21-28
• /
• 1998

This study develops a variable structure controller using the time-varying nonlinear sliding surface instead of the fixed sliding surface, which has been the robustness against parameter variations and extraneous disturbance during the reaching phase. By appling TS fuzzy algorithm to the regulation of the rionlinear sliding surface, the reaching time of the system trajectory is faster than the fixed method. This proposed scheme has better performance than the conventional method in reaching time, parameter variation and extraneous disturbance. To demonstrate its performance, the proposed control algorithm is applied to a rotational inverted pendulum.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.6 no.3
• /
• pp.40-50
• /
• 1996

By using the neural networks(NN) as system identifier, the on-line self tuning method for fuzzy controller(FC) is proposed. In theis method, the learning of NN is carried out during control operation of FC and the cinsequent parameters of FC is tuned on-line automatically by means of system output errors backpropagated through NN. The Sugeno fuzzy model with constants as consequent parameters is selected for simplifying computation. In procedures of parameter tuning, the gradient descent method is used and the gradient vectors for adjusting the weight of NN are transferred as controller output errors. To evaluate the performance, the proposed method is applied to the inverted pendulum system.

• Journal of Advanced Marine Engineering and Technology
• /
• v.32 no.5
• /
• pp.785-790
• /
• 2008

A technique for determining the feedback gain of the states feedback controller using a real-coded genetic algorithm(RCGA) is presented. It is concerned with the states error to the performance index of a RCGA. As for assessing the performance of the controller three performance criteria (ISE. IAE and ITAE) are adopted. And designing the controller involves a constrained optimization problem. Therefore a real-coded genetic algorithm incorporating the penalty strategy is used. The performance of the proposed method is demonstrated through a set of simulation about an inverted pendulum system.