All issues
Volume 44, Issue 9
Displaying 1-9 of 9 articles from this issue
- |<
- <
- 1
- >
- >|
-
Mitsuru TAKAHASHI, Manabu GOUKO, Koji ITO2008 Volume 44 Issue 9 Pages 699-704
Published: September 30, 2008
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSEvent Related Desynchronization (ERD) which is a specific brain wave of motor imagery is popular in the Brain Computer Interface (BCI) researches. Several researches showed that sensory feedbacks affected ERD, but it is not clear which sensory imputs (tactile, muscular, tendonious, articular) mainly affect ERD. We examined how ERD is affected by the Functional Electrical Stimulation (FES) on both feet for healthy subjects. The resuits indicated that bigger ERD was extracted as the FES stimuls increase. And bigger ERD was also extracted during the leg free condition than the leg fixed condition. It suggests that the muscular and articular sensations induce ERD on foot motor area (Cz).View full abstractDownload PDF (12139K) -
On the Relation between Synthesis Conditions Based on a Domain of L2/ExponentialStabilit yPerformanceTsuyoshi KIYAMA2008 Volume 44 Issue 9 Pages 705-713
Published: September 30, 2008
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSThis paper considers linear time-invariant continuous-time control systems with saturation and/or dead-zone nonlinearities, and proposes two synthesis methods satisfying a regional L2 and/or exponential stability performance for the system based on the generalized sector approach. One of them is an integrated design of dynamical output feedback and anti-windup controllers, while the other is an integrated design of static output and state feedback controllers. The two methods assume the output of the nonlinearities to beavailable for the control. In this case, this paper indicates that the synthesis problems using the two methods can be recast as linear matrix inequality (LMI) optimization problems, respectively. Furthermore, it is proved that the synthesis condition based on the former dynamical controller is a sufficient condition of the corresponding one based on the latter static controller. In the special case of the same class of the two controller input signals, it is also proved that two achievable control performances using the two controllers are exactly the same. In particular, it is clarified that the synthesis condition using the dynamical controller is composed of a synthesis condition of a state feedback gain for control systems with saturation and/or dead-zone nonlinearities, and a synthesis condition of an observer gain for linear control systems without the nonlinearities. Finally, this paper points out that the proposed synthesis method is helpful through numerical examples designing control systems with anti-windup control structures via the derived LMI conditions.View full abstractDownload PDF (5240K) -
Masato ISHIKAWA2008 Volume 44 Issue 9 Pages 714-720
Published: September 30, 2008
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSThis paper is concerned with stabilizing feedback control for a class of nonholonomic driftless systems, whose controllability Lie algebra rank condition is satisfied by up to first-order Lie brackets. We propose a switched feedback law which drives all the initial states to the origin with bounded control inputs (as opposed to unbounded, division-by-zero-type discontinuous control). The discontinuity of the feedback law takes place on a subspace defined by the ‘parallelism’ condition for the base and the fiber vectors in R3 (or simply q×→=0). We also show that the complement set of this discontinuity region is homotopic to SO (3) which is isomorphic to an SO (2)-bundle on S2. The proposed control law is examined by numerical simulations.View full abstractDownload PDF (4516K) -
Naoya HATAKEYAMA, Akira SHIMADA2008 Volume 44 Issue 9 Pages 721-728
Published: September 30, 2008
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSWe propose a high-speed robust motion control technique for inverted pendulum robots that utilizes forward and backward tilting as a control factor. An inverted pendulum is a self-regulated system simulating a child's game of swaying an umbrella or a stick upwards. The controller design for pendulums has been widely challenged since the1970s. A two-wheeled, self-balancing electric transportation device using the inverted pendulum's control principle was developed in the US. Many biped walking robots have also made use of this. principle. Furthermore, the feature of space-saving of inverted pendulum robots will be highly regarded and they will contribute to our better lives. On the other hand, essentially, inverted pendulums possess the control capability that they do not topple down. Shimada and Hatakeyama suggested an idea that was contrary to this basic principle. Their controller was designed to brake down its balance when in motion. This was done using zero dynamics derived by partial feedback linearization, in order to control revolving and curving motion. However, this control system is a feed-forward control system making use of forward and backward tilting and it is not robust. To solve this problem, we have designed a two degrees of freedom controller based on the feed-forward controller and H∞ control technique. Finally we present the simulation and experimental results for validity.View full abstractDownload PDF (12228K) -
Jin-Hua SHE, Hironori KOBAYASHI, Xin XIN, Atsushi IMAZU, Yasuhiro OHYA ...2008 Volume 44 Issue 9 Pages 729-734
Published: September 30, 2008
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSIn this paper, a new method of planning the swing-up motion of an acrobot is presented based on its fallingdown motion. First, the initial state of the acrobot is set to the unstable upright equilibrium position, and the movement from this state to the stable downward equilibrium position is produced by introducing an artificial damping torque. Next, the motion planning problem is formulated as a constrained nonlinear optimal problem, and an optimal damping coefficient is obtained by solving this problem. Then, the swing-up motion is taken to be the reverse of the falling-down motion. The validity of this method is demonstrated through simulations.View full abstractDownload PDF (4639K) -
Application of 2-Dimensional Adaptive ControlYoshihiko MIYASATO2008 Volume 44 Issue 9 Pages 735-742
Published: September 30, 2008
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSThis paper provides an alternative approach to solve iterative learning control (ILC) of robotic manipulators by introducing hybrid adaptation schemes and extended versions of those. The hybrid adaptation schemes are adaptive control structures which involve continuous-time control of processes and discrete-time updates of tuning parameters simultaneously. The various hybrid adaptation schemes are summarized, and the extended versions of those, that is, “2-dimensional adaptive control schemes”, are also proposed. 2-dimensional adaptive control strategies contain2types of adaptation processes, and provide more skillful learning properties where adaptive processes themselves can be improved adaptively.View full abstractDownload PDF (4699K) -
Control under Constrained Input and an Application to the SOA AnalysisTsuyoshi SAGAMI, Norikazu ITO, Shigeki NAKAURA, Mitsuji SAMPEI2008 Volume 44 Issue 9 Pages 743-750
Published: September 30, 2008
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSIn this paper, a design of the discontinuous controller for the chained system is introduced.The proposed method is based on the the Time State Control Form (TSCF) approach.Firstly, chained form system is converted into a set of TSCF's through a discontinuous time-scale transformation.The proposed representation includes Astolfi's discontinuous controller as its special case, and thus enables us to relax the state regulation condition.Two application examples are presented.The first one concerns the problem of a controller modification under input constraints.The Astolfi controller is modified so that it will not violate the given constraints without narrowing its set of attraction.State convergence is assured by the relaxed regulation condition derived from the proposed TSCF representation.The second one is the enlargement of the Set Of Attraction (SOA) through the design of the time scale function.Firstly, nonlinear SOA analysis problem is transformed to the linear one.Then the time scale function is shaped so that the inner approximation of the SOA grows as large as possible under the given state constraints.View full abstractDownload PDF (8378K) -
Kunihiko HIRAISHI, Koichi KOBAYASHI, Sunseong CHOE2008 Volume 44 Issue 9 Pages 751-759
Published: September 30, 2008
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSA notion of bounded bisimilarity is introduced for discrete-time piecewise linear systems. Bounded bisimilarity is the same as the usual bisimilarity except that only trajectories of finite length are considered. We can always find a finite set of equvalence classes for bounded bisimilarity. Using bounded bisimilarity, we propose a method for computing control inputs that achieve a state trajectory satisfing given temporal logic formula and also a desirable level for a given objective function.View full abstractDownload PDF (10061K) -
Tatsuro BABA, Nobuo OHMAE, Koichi OSUKA2008 Volume 44 Issue 9 Pages 760-765
Published: September 30, 2008
Released on J-STAGE: February 25, 2013
JOURNAL FREE ACCESSWe developed a technique to control the velocity range and offset of ultrasound Doppler images based on the histogram of Doppler auto-trace waveforms for the purpose of easy operation in clinical use. Conventional techniques are un-stable in the presence of poor Doppler signal or aliasing. We developed a new system based on hybrid control. We compared the performances with the conventional system by simulation. The stable response to various state changes assumed in normal clinical application was obtained.View full abstractDownload PDF (23146K)
- |<
- <
- 1
- >
- >|