All issues
Volume 39, Issue 7
Displaying 1-14 of 14 articles from this issue
- |<
- <
- 1
- >
- >|
-
2nd Report, Complex Multi-Resolution Analysis with RI-Spline WaveletZhong ZHANG, Hiroshi TODA, Hiroaki KAWABATA2003 Volume 39 Issue 7 Pages 615-623
Published: July 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSWavelet transform, which is a well-known and a remarkable way for the unsteady signal processing, is considered to surpass the Fourier analysis and it is applied in various fields. In this study, we propose a complex multi resolution analysis (CMRA) with RI-Spline wavelet and dual-tree algorithm corresponding to CMRA in order to carry out discrete wavelet transform, and apply it to remove noise. The main results obtained can be summarized as follows: 1) Poor shift invariance which occurred by the discrete wavelet transform in the case of using real mother wavelet has been improved by using CMRA was conformed. 2) The discrete wavelet transform corresponding to CMRA has been applied to removal of the white noise in ECG and the pulse noise in record music, and the usefulness of our approach was demonstrated.View full abstractDownload PDF (1973K) -
1st Report: -Verification of Repeatability, Reproducibility and Zero Error-Koji OHGUSHI, Takashi OTA, Kazunaga UEDA2003 Volume 39 Issue 7 Pages 624-630
Published: July 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSAs uncertainty contributions for the static torque calibration of a torque measuring device, the following components are considered; repeatability with unchanged mounting position, reproducibility with changed mounting position, zero error, interpolation error, reversibility (hysteresis) and resolution including the ability of the indicator. Some foreign standard documentations, however, have not revealed any experimental confirmation for the validity of assumed probability distribution for each contribution. In order to establish a determination method of uncertainty for the torque calibration, the validity of probability distributions was verified by the numerous repetitions of torque measurement using the primary torque standard machine. The determination method is proposed based on the experimental results.View full abstractDownload PDF (2559K) -
2nd Report: -Verification of Interpolation Error, Reversibility and Resolution-Koji OHGUSHI, Takashi OTA, Kazunaga UEDA2003 Volume 39 Issue 7 Pages 631-636
Published: July 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn order to establish a determination method of uncertainty for the static torque calibration, the validity of probability distributions was verified by the numerous repetition of torque measurement using the primary torque standard machine. In the 1st report, assumed probability distributions were investigated for repeatability with unchanged mounting position, reproducibility with changed mounting position and zero error. This paper describes results of experimental confirmation for other uncertainty contributions: interpolation error, hysteresis (reversibility) and resolution including the ability of the indicator. Total uncertainty for calibration of the torque measuring device is also calculated and the determination method is proposed.View full abstractDownload PDF (2369K) -
Shigenori SANO, Kazuya OGATA, Yoshikazu HAYAKAWA, Shoji TAKAGI2003 Volume 39 Issue 7 Pages 637-645
Published: July 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSFor the control design, it is important to make a good model. To identify the model parameter exactly, it is necessary to use data which are less affected by the disturbance or modelling uncertainty. We have proposed that it is important to evaluate the upper bound of the parameter estimation error when the disturbance and the unstructured model uncertainty exist. We design the input signal and the frequency weight for least square criterion such that the upper bound is minimized and confirm that it is useful to identify using the input signal and frequency weight. In this report, the input signal is limited to the periodic signal.View full abstractDownload PDF (1897K) -
Ryuji ENOMOTO, Masasuke SHIMA2003 Volume 39 Issue 7 Pages 646-653
Published: July 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe paper deals with the global asymptotic stabilization problem for time-invariant general nonlinear state equations from the viewpoint of the topological classification of state feedback controlled systems. A notion of “Boundary tangency manifolds of the state equation” can be derived from the generalized Poincaré-Hopf index formula by C.C. Pugh and C. McCord. We discuss the relation between the boundary tangency manifolds of the state equation and the index formula on the input manifold which is a natural geometrical object for the differentiable state feedback control law. We also introduce terms of the dissipative boundary and the neutral boundary of the state equation and discuss the elementary results for the global asymptotic stabilization problem.View full abstractDownload PDF (2096K) -
Ryuji ENOMOTO, Masasuke SHIMA2003 Volume 39 Issue 7 Pages 654-661
Published: July 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper presents a new approach to the classification problem of topological structures of feedback controlled systems for time-invariant general nonlinear state equations. The generalized Poincaré-Hopf index theorem by C.C. Pugh and C. McCord leads us to a notion of “Boundary tangency manifolds of the state equation”. We show a topological method to judge whether the controlled system is circumscribed or inscribed at a point of the boundary of a compact and connected submanifold in the state space. For this purpose, we discuss the intersection theory for boundary tangency manifolds and an input manifold which is the geometrical object for a differentiable state feedback control law.View full abstractDownload PDF (2274K) -
Keitaro OHNO, Roberto HOROWITZ2003 Volume 39 Issue 7 Pages 662-669
Published: July 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, we propose a new multi-rate state estimator, which has a proportional, integral, and discontinuous error feedback structure to improve robustness. Conventional linear state estimators do not have enough accuracy under the existence of external disturbance and model uncertainty. With our proposed estimator, output tracking performance can have better robustness. We mainly focus its application on a system that has a hardware difficulty of obtaining high enough measurement sampling frequency, e.g., Hard Disk Drives. Integral and discontinuous actions are introduced to alleviate the estimation error caused by the disturbance and model uncertainty, while multi-rate technique is to obtain a smoother control input and higher bandwidth. Integral action and multi-rate technique also work as reducing the chattering motion caused by the discontinuous action. Simulations and experiments have been carried out to validate our proposed multi-rate state estimator. Finally, an experimental example of seeking control for a single stage actuated hard disk drive is demonstrated.View full abstractDownload PDF (2073K) -
Ikuko NISHIKAWA, Tomoyuki NAKANISHI2003 Volume 39 Issue 7 Pages 670-678
Published: July 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA statistical method is proposed to cope with a large number of discrete states in a given state space in reinforcement learning. As a coarse-graining of a large number of states, less number of sets of states are defined as a group of neighboring states. State sets partly overlap each other, and one state is included in a multiple sets. The learning is based on an action-value function for each state set, and an action-value function on an individual state is derived by a statistical average of multiple value functions on state sets at the time of an action selection. The proposed method is applied to a board game Dots-and-Boxes. The state sets are defined as subspace templates of a whole board state with dots and lines, taking a geometric symmetry into the consideration. A reward is given as a number of acquired boxes minus lost boxes. Computer simulations show a successful learning through the training games competing with a mini-max method of the search depth 2 to 5, and the winning rate against a depth-3 mini-max attains about 80%. An action-value function derived by a weighted average with the weight given by the variance of rewards shows the advantage compared with an action-value function derived by a simple average.View full abstractDownload PDF (1839K) -
Jun IZAWA, Toshiyuki KONDO, Koji ITO2003 Volume 39 Issue 7 Pages 679-687
Published: July 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe present paper proposes a learning control method for the musculoskeletal system of arm based on the reinforcement. An optimization for the hand trajectory and muscle's force distribution is needed to acquire the reaching motion. The proposed architecture can acquire an optimized motion through learning the task. However, the biological control system composed of muscluloskeletal system is not able to sense the state without time delay. The time delay causes instability of learning. The proposed scheme consists of reinforcement part and neural internal model. Neural internal model is employed to compensate for the time delay. Then, there must be a modeling error if the muscle noise is assumed. Thus we introduce the minimum modeling error criterion for reinforcement learning. The minimum modeling error criterion gives not only reduction of total muscle level but also smoothness of the hand trajectory. The effectiveness and the biological plausibility of the present model is demonstrated by several simulations.View full abstractDownload PDF (2493K) -
Hiroshi HARADA, Hiroshi KASHIWAGI, Toshiyuki ANDOH, Kazuki KABA2003 Volume 39 Issue 7 Pages 688-690
Published: July 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThe authors have recently proposed a new signal processing technique called M-transform. In this paper, the authors propose a new method for impulsive noise reduction by use of M-transform.View full abstractDownload PDF (396K) -
Kazuaki MASUDA, Eitaro AIYOSHI2003 Volume 39 Issue 7 Pages 691-693
Published: July 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSWe propose a new dynamics for solving optimization problems constrained in the internal space of a simplex, whose discretization generates chaos. The efficiency of the optimization method which anneals the chaotic trajectories is shown with numerical simulations.View full abstractDownload PDF (429K) -
Toshiaki SUZUKI, Toshiyuki TANAKA2003 Volume 39 Issue 7 Pages 694-696
Published: July 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn inverse-GPS (Global Positioning System), the intersymbol interference greatly influences on the positioning error. This paper deals with new interference canceller that reduces the error. Our proposed interference canceller consists of DLL (Delay Locked Loop) and adaptive filter. Only a primary phase of replica signal is estimated in DLL. Afterwards, the strict wave is generated with the adaptive filter. We could decrease the intersymbol interference error caused from the user increase, using interference canceller that combined DLL with the adaptive filter.View full abstractDownload PDF (441K) -
Weishui WAN, Kotaro HIRASAWA, Jinglu HU2003 Volume 39 Issue 7 Pages 697-698
Published: July 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper a new algorithm applying Gram-Schmidt orthogonalization algorithm to the outputs of nodes in the hidden layers is proposed with the aim to reduce the interference among the nodes in the hidden layers, therefore to enhance the generalization ability of neural networks, which is much more efficient than other regularizers methods. Simulation results confirm the above assertion.View full abstractDownload PDF (343K) -
Kosuke UMESAKO, Masanao OBAYASHI, Kunikazu KOBAYASHI2003 Volume 39 Issue 7 Pages 699-701
Published: July 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSWe propose a reinforcement learning system that is constructed autonomously according to the state of environment observed during trials and the efficient method of selecting action whose direction varies. Effectiveness is shown through mobile robot control.View full abstractDownload PDF (556K)
- |<
- <
- 1
- >
- >|