-
[show abstract]
[hide abstract]
ABSTRACT: This paper describes a control strategy for throwing motion of the springed Pendubot based on the concept of unstable zero dynamics. An underactuated two-link planar robot called the Pendubot is investigated to realize dexterous actions of the superior limb in human throwing motion. A torsion spring is mounted on the passive joint of the Pendubot representing the flexibility of the cubital joint. In the proposed control strategy, the zero dynamics is intentionally destabilized when the end-effector of the springed Pendubot is constrained on a geometric path via output zeroing control for the deviation between the end-effector and the geometric path. The unstable zero dynamics drives the end-effector along the geometric path to achieve a fast and accurate throw in a desired direction when the input is devoted to constrain the end-effector on the geometric path. The unstable zero dynamics is analytically derived to guarantee the divergence of the end-effector along the geometric path. Numerical simulations confirm the effectiveness of the proposed control strategy.
Control Applications (CCA), 2010 IEEE International Conference on; 10/2010
-
[show abstract]
[hide abstract]
ABSTRACT: Inverse optimal design for bilinear systems is considered. The main result is that a nonlinear optimal feedback control law which minimizes a new quadratic cost function with nonlinear weight is obtained based on an inverse optimal control problem for bilinear systems. This inverse optimal control design is applied to the problem of the stabilization of the inverted pendulum on the cart which moves not only in the horizontal direction but also in the vertical direction. This inverted pendulum system can be transformed into a bilinear system by using input transformation and coordinate transformation focused on the center of percussion of the pendulum. It is theoretically shown that the proposed nonlinear optimal feedback controller has higher control performance than a conventional linear optimal controller for the linear approximation system. Furthermore, it is shown by numerical simulations that the control performance of the pendulum is improved by utilizing the vertical movement of the pendulum.
Decision and Control, 2009 held jointly with the 2009 28th Chinese Control Conference. CDC/CCC 2009. Proceedings of the 48th IEEE Conference on; 01/2010
-
[show abstract]
[hide abstract]
ABSTRACT: This paper describes fly-casting motion control for a multilinked manipulator named `fly-casting manipulator', which consists of a casting rod, a fly line and a sinker. This manipulator simulates fly-fishing rod and has flexible potential. Flexible structure such as fly-fishing rod has advantages of accumulating energy because of its flexibility. Fly-casting motion has three phases, back-casting motion, collision and forward casting motion. Back-casting motion and forward-casting motion are energy accumulating phases and collision is energy losing phase. Additionally, the number of control inputs is less than the degree of freedom of the fly-casting manipulator. Thus, it is not possible to control all the states. Since this property makes it difficult to handle the casting rod, control problem of fly-casting motion is interesting as a point of view of control theory. Fly-casting manipulator is modeled as a multilinked underactuated manipulator. Only base joint is active joint, while the others are passive and constrained by torsion springs. The springs generate bending force which makes manipulator control easier. The proposed controller for casting motion is based on an input-output linearization and an output zeroing control to constrain the sinker on the desired path. Simulation results show the effectiveness of the proposed controller.
Decision and Control, 2009 held jointly with the 2009 28th Chinese Control Conference. CDC/CCC 2009. Proceedings of the 48th IEEE Conference on; 01/2010
-
[show abstract]
[hide abstract]
ABSTRACT: This paper deals with motion control of throwing generated by dexterous action. Dexterous actions can be seen in many sports. In baseball pitching, dexterous throwing seems to use energy transfer and a physical constraint at the elbow joint. To implement the dexterous throwing, two types of two-link underactuated manipulator are presented. One model has a spring at 2nd joint which represents an arm's stiffness and a constraint at elbow joint, another model has a physical absolute constraint at elbow joint. For these models, throwing motion control method based on output zeroing which specifies the path of the ball held by an end-effector is proposed. The proposed control strategy realizes the energy efficient motion for throwing to the desired direction. Simulation and experimental results show the effectiveness of the proposed control method.
American Control Conference, 2009. ACC '09.; 07/2009
-
[show abstract]
[hide abstract]
ABSTRACT: This paper describes the modeling of the Acrobot composed of 2 links with a curved contour and the continuous rolling motion control for this Acrobot. The outer contour of each link is shaped from the arc whose radius is different by the tip and the sidepiece of the link. The model differs according to the contact point between the Acrobot and the ground. Therefore, it is difficult to control the whole motion via common control strategy. From an intuitive analysis based on the Acrobot¿s energy while the Acrobot rolls with a certain constant relative angle, control strategy is constructed in two phases. The phases are when the Acrobot rolls with lowering( downward phase) and raising(upward phase) the center of the gravity. As the control in upward phase needs to lower the COG of the Acrobot, a collision between the Acrobot and the ground becomes unavoidable. Therefore, analysis of this collision phenomenon is also conducted. At last, by selecting the output functions that can achieve the control objective at each phase and applying Output Zeroing control, continuous rolling motion is realized in numerical simulation.
Decision and Control, 2008. CDC 2008. 47th IEEE Conference on; 01/2009
-
[show abstract]
[hide abstract]
ABSTRACT: Recently, nonlinear H<sub>¿</sub> control theory has been paid attention. The solvable condition of nonlinear H<sub>¿</sub> control problem is given by the Hamilton Jacobi inequality (HJI). State-Dependent Riccati Inequality (SDRI) is one of approaches to solve the HJI. The SDRI contains state-dependent coefficient (SDC) form of a nonlinear system. The SDC form is not unique, so free parameters of it is considered. If bad SDC form is chosen, then there is no solution of SDRI. In this paper, the relationship between free parameters and SDRI is clarified. The free parameters are generated when SDRI is derived from HJI. And they affect the conservativeness of SDRI. Then new method of design free parameters which reduces the conservativeness of SDRI is proposed. Finally, numerical examples to verify the effect of this method is shown.
Decision and Control, 2008. CDC 2008. 47th IEEE Conference on; 01/2009
-
[show abstract]
[hide abstract]
ABSTRACT: This paper describes a control structure for planar humanoid runner on uneven terrain utilizing output zeroing. In this paper, the terrain whose inclination changes at random within a certain range is called uneven terrain. First, the change of dynamics by running on uneven terrain is considered, and the change of dynamics by running on uneven terrain is negated by adding the impulse input. The running on uneven terrain whose inclination changes within plusmn7 [deg] is achieved by the simulation. Second, important parameters of the output function is modified online utilizing Q-learning to acquire the running motion which can dynamically cope with the inclination of terrain. As a result, the running on uneven terrain whose inclination changes gradually within plusmn5[deg] is achieved.
SICE Annual Conference, 2008; 09/2008
-
[show abstract]
[hide abstract]
ABSTRACT: This paper describes the throwing motion control experiment utilizing a 2 link arm with a passive joint. In throwing, dexterous actions can be seen. Dexterous throwing uses energy transfer from a trunk of the body to a hand and a physical constraint at the elbow joint. However, human throwing is very difficult to analyze all features, because throwing motion performs in three dimensions. Therefore, under-actuated 2 link model is applied in this paper. Proposed model has a spring joint which is represented as an armpsilas stiffness and a constraint at elbow joint. For this model, the control method based on output zeroing which specifies the path of the ball at the end-effector is proposed. By developing an experimental equipment and carrying out experiments of throwing motion control, the effectiveness of the proposed control method is verified.
SICE Annual Conference, 2008; 09/2008
-
[show abstract]
[hide abstract]
ABSTRACT: This paper describes swing up control for the acrobot with the compliance of a high bar based on the energy pumping method. The energy interaction exists between a gymnast and the high bar, because the high bar is bending during gymnastics. Therefore, more efficient motion would be constructed by controlling this energy interaction. In this paper, swing up control which considers the compliance and the energy interaction with each component is proposed, and the advantage and the disadvantage by using compliance are examined by comparing numerical simulations for with and without compliance cases.
Decision and Control, 2007 46th IEEE Conference on; 01/2008
-
[show abstract]
[hide abstract]
ABSTRACT: This paper describes dynamic position control of surface vessel with disturbances such as waves and wind. A smooth time-varying controller is proposed that performs both in presence and absence of disturbances. Using this controller, position of the vessel converges to neighborhood about zero that can be made arbitrarily small. In addition, if disturbance is small, orientation of the vessel converges to neighborhood about zero that is as small as possible. Effectiveness of proposed control law is demonstrated by simulation.
Decision and Control, 2007 46th IEEE Conference on; 01/2008
-
[show abstract]
[hide abstract]
ABSTRACT: This paper describes a control structure for humanoid runner which is using Q-learning and output zeroing. Based on this structure, the sagittal plane motion control algorithm is formulated using output zeroing. And Q-learning algorithm is used to learn key parameters of output function so that humanoid robot can achieve desired running motion. For the parameters of output function to be learned, the parameters which control the hip trajectory and torso angle of humanoid are selected. For an evaluation of learning, the running speed and stride of humanoid are used. As a result of learning, it is confirmed that key parameters of output function converge on constant value and that humanoid robot can realize continuous running in numerical simulation. Besides it is also verified that humanoid robot can realize continuous running at uneven terrain using similar control structure.
Decision and Control, 2007 46th IEEE Conference on; 01/2008
-
[show abstract]
[hide abstract]
ABSTRACT: This paper describes an experiment for swing up control for the acrobot with the compliance of the high bar based on the energy pumping method. The energy interaction exists between the gymnast and the high bar, because the high bar is bending during gymnastics. Therefore, more efficient motion would be constructed by controlling this energy interaction. Based on this probability, swing up control which considers the compliance and the energy interaction with each component is proposed. To confirm this discussion, an experiment using the acrobot is carried out, and the advantage and the disadvantage by using compliance are examined by comparing experimental results for with and without compliance case.
Intelligent Robots and Systems, 2007. IROS 2007. IEEE/RSJ International Conference on; 12/2007
-
[show abstract]
[hide abstract]
ABSTRACT: For the bipedal robot, this paper investigates natural running motion like human running, without performing tracking control to the optimal trajectory of such a motion. The bipedal robot considered in this paper is planar and has revolute joints which actuate ankle, knee and hip. The running motion is assumed to consist of two successive phases which are composed of a flight phase and a stance phase: the flight phase where the bipedal robot floats, the stance phase where either a heel or a toe keeps contact with the ground. In order to realize the natural motion of the human, an output function which is easily understood in the physical aspect is designed on the basis of the human actual motion. Running motion of bipedal robot is generated autonomously by zero dynamics which appears by zeroing output function. Some parameters of this output function are re-configured online in order to suppress disturbance caused by unstable zero dynamics. Through a numerical simulation it is confirmed that running motion becomes stable
Decision and Control, 2006 45th IEEE Conference on; 01/2007
-
[show abstract]
[hide abstract]
ABSTRACT: An analysis of a cascade control system constructed by an inner and an outer feedback loop is presented. It differs from other researches, because the inner feedback loop is unstable. The control method is motivated by balance stabilization of a humanoid robot based on the zero moment point (ZMP) principle where the effectiveness is well-known experimentally. This paper generalizes the control method, and analyzes the stability and the robustness w.r.t. parameter variations of the plant and external disturbances. Some examples to the ZMP control and a simple linear system are described and the simulation results are shown
Computer Aided Control System Design, 2006 IEEE International Conference on Control Applications, 2006 IEEE International Symposium on Intelligent Control, 2006 IEEE; 11/2006
-
[show abstract]
[hide abstract]
ABSTRACT: Most of the inputs of actual plants have the limitations of variation range. These limitations may be due to physical restrictions on the actuators. It is well known that the effects of actuator saturation cause the deterioration of control performance called windup phenomena. In this paper, a novel anti-windup strategy for multi-input multi-output (MIMO) systems subject to actuator saturation is proposed. The method is based on the notion of timescale transform. Some simulation and experimental results are presented to show the effectiveness of the proposed method
Computer Aided Control System Design, 2006 IEEE International Conference on Control Applications, 2006 IEEE International Symposium on Intelligent Control, 2006 IEEE; 11/2006
-
[show abstract]
[hide abstract]
ABSTRACT: This paper deals with sinus-lifting motion which is some animal snake motion such as raising part of body. This motion is known to be the effective action in dynamics and anatomy fields. Our approach in this paper is to analyzed this motion in the sense of control theory. A snake-like robot which consists of active joints and passive wheels with no side-slip condition is used in this paper. The concept of dynamic manipulability of a snake-like robot is applied to the general locomotion control for this model. Numerical simulations based on this manipulability show that sinus-lifting motion improve the manipulability and affect the period of body shape.
SICE 2004 Annual Conference; 09/2004
-
[show abstract]
[hide abstract]
ABSTRACT: Devil stick is a kind of juggling as to control a floating stick (the center stick) by hitting or pushing with other sticks (the hand sticks). Various stick motions are performed by jugglers and some of them are studied as control problem. The purpose of this paper is to rotate the center stick by only pushing with one hand stick. In other word, control objective is to maintain a position and an angular velocity of the center stick. The control strategy in this paper is as follows. To begin with, some states are controlled by output zeroing control. The output function is derived from observing the motion of good jugglers. But there are two problems in this strategy. First, since an attitude angle of the center stick is in unobservable subspace, the angular velocity can't be controlled. Secondly, a direction of the input force isn't considered. For the first problem, by solving equations of the zero dynamics, it turns out that it is possible to stabilize the angular velocity by changing a contact point of the center stick and the hand stick. For the second one, the analytical result shows that the required force depends on the angular velocity. Some numerical simulations show the usefulness of the proposed strategy.
Decision and Control, 2003. Proceedings. 42nd IEEE Conference on; 01/2004
-
Proc. of the 6th IFAC NOLCOS Symposium; 01/2004
-
[show abstract]
[hide abstract]
ABSTRACT: This paper deals with hopping motion analysis of spherical robot with a typical impact condition like 'superball'. A horizontal velocity of super ball after an impact depends on the angular velocity before the impact where it doesn't slip at the contact point. Our robot model consists of an elastic spherical shell with such an impact condition and of an internal rotary input. Assuming the perfectly elastic impact in a vertical direction, a feedback control strategy for generating the horizontal velocity after an impact by the rotary input during flying phase is derived. Simulation results show that some hopping motion is achieved by the proposed method. In addition, this paper shows the possibility and necessary condition that a climbing motion with impacts against two parallel vertical planes is achieved.
Intelligent Robots and Systems, 2003. (IROS 2003). Proceedings. 2003 IEEE/RSJ International Conference on; 11/2003
-
[show abstract]
[hide abstract]
ABSTRACT: A large number of studies about a swing up control on a horizontal bar have been made, but compliance of a horizontal bar is ignored in their studies. This paper deals with the swing up control using compliance, and shows that proposed control law and the way of selecting physical parameters of a horizontal bar in consideration of compliance are effective. A feedback control to swing up by considering the energy of a link and the optimal trajectory to increase total mechanical energy of a link for one expansion motion is proposed. Advantage to use compliance is confirmed through numerical simulation.
SICE 2003 Annual Conference; 09/2003
