巻号一覧
42 巻, 12 号
選択された号の論文の11件中1～11を表示しています
- |<
- <
- 1
- >
- >|
-
青野 俊宏, 小渡 武彦2006 年 42 巻 12 号 p. 1273-1282
発行日: 2006/12/31
公開日: 2009/03/27
ジャーナル フリーThe purpose of this research is to develope a method of back flow compensation using frequency characteristics of the signal from an air flow sensor (AFS). Parameters of the AFS signal which has strong correlation with back flow ratio were searched. The parameters were searched on terms of simulation, mathematical analysis and experiments.
The simulation results showed that normalized powers f2/f1 and f3/f1 have strong correlations with back flow ratio, where f1 is the intensity of the frequency of the pulsation, f2 is the intensity of twice of the frequency of the pulsation, and f3 is the intensity of three times of the frequency of the pulsation. These two correlations were robust to the design of the induction system, engine revolution speed and load. The back flow ratio can be estimated using normalized power f2/f1 and f3/f1 within the accuracy of 5%. The correlations between the normalized powers f2/f1, f3/f1 and the back flow ratio were validated by mathematical analysis. The correlations were also evaluated by experiments using four cycle engine. The experimental results showed that the back flow ratio was estimated using the normalized powers f2/f1, f3/f1 within the accuracy of 5%.抄録全体を表示PDF形式でダウンロード (2807K) -
佐久間 史洋, 馬 莱娜2006 年 42 巻 12 号 p. 1283-1287
発行日: 2006/12/31
公開日: 2009/03/27
ジャーナル フリー0.65μm radiation thermometers are used as the standard thermometers at high temperatures. The spectral stability of these thermometers influences largely on the calibration uncertainty. Therefore it is important to check the spectral stability. Usually spectral stability was estimated by using monochromators. But it is difficult for a calibration laboratory to maintain a monochromator to measure the spectral characteristics of the radiation thermometers. We propose to use two fixed-point calibrations to check the spectral stability. We compared the effective wavelengths obtained from the monochromator and those from the fixed point ratios. For the fixed points, copper and silver were employed. Unfortunately the silver point is almost the lower limit of the 0.65μm radiation thermometers. Resolution and uncertainty in determining the output ratio between copper and silver points were contributed mainly from the short-term and long-term stability of fixed points which were about 0.02% and 0.1% of the radiation thermometer output. These values correspond to 0.08nm resolution and 0.6nm uncertainty for the effective wavelength estimation.抄録全体を表示PDF形式でダウンロード (1106K) -
田中 良幸, 松下 和寛, 辻 敏夫2006 年 42 巻 12 号 p. 1288-1294
発行日: 2006/12/31
公開日: 2009/03/27
ジャーナル フリーThe present paper develops a virtual curling system using robotic devices based on the impedance training, with which a trainee manipulates the handle of the impedance-controlled robot and slides the virtual stone toward the center of the house, as similar as the actual curing game. The developed system can estimate trainee's hand impedance during the curing task to quantitatively evaluate an adaptation ability of musculoskeletal characteristics of the upper extremity. To design a training program and an evaluation index of training effects, standard hand motion and its dynamic properties for the virtual curling task are clarified through a set of training experiments with 18 volunteers.抄録全体を表示PDF形式でダウンロード (2356K) -
水本 郁朗, 大平 聡, 公文 誠, 岩井 善太2006 年 42 巻 12 号 p. 1295-1304
発行日: 2006/12/31
公開日: 2009/03/27
ジャーナル フリーThis paper deals with a design problem of model output following control system for non-uniformly sampled systems with a non-uniform fast-rate sampling of the input and slow-rate sampling of the output. The proposed method attains the model output following by using CGT (Command Generator Tracker) input of a fast-rate and ensures the stability of the control system by the fast-rate output feedback based on an almost strictly positive realness (ASPR-ness) of the controlled system with an estimated output and a parallel feedforward compensater (PFC).抄録全体を表示PDF形式でダウンロード (2080K) -
桜間 一徳, 中野 和司2006 年 42 巻 12 号 p. 1305-1312
発行日: 2006/12/31
公開日: 2009/03/27
ジャーナル フリーThis paper addresses online modification of reference trajectories for multiple robots with collision avoidance. First, a smooth mapping dependent on the relative distances between the robots is proposed to coordinate collision-free motion from given reference trajectories. The trajectories modified by the mapping make the robots avoid collisions with each other, and are smooth enough to be tracked by the robots. Moreover, when the reference trajectories are separated from each other, the mapping is identical and the modified trajectories remain unchanged from the original ones. Next, a value function is considered consisting of the distances between the original and modified reference trajectories. From the viewpoint of minimizing the value function, a way to choose design parameters for the mapping is specified. Finally, a simulation for three mobile robots demonstrates the effectiveness of the proposed method.抄録全体を表示PDF形式でダウンロード (1989K) -
新田 益大, 杉本 謙二2006 年 42 巻 12 号 p. 1313-1319
発行日: 2006/12/31
公開日: 2009/03/27
ジャーナル フリーThis paper proposes a method for identifying and suppressing unknown disturbances based on independent component analysis. Under the assumption that the sources of disturbance and control inputs are mutually independent, the proposed method makes it possible to estimate the transfer functions from disturbance sources to signals to be controlled. Two types of disturbance suppression control systems using the disturbance model are introduced and the effectiveness of these control systems has been demonstrated through some numerical simulations.抄録全体を表示PDF形式でダウンロード (1482K) -
Computation of an Output Admissible SetMizuyo TAKAMATSU, Yasuaki OISHI, Kenji HIRATA2006 年 42 巻 12 号 p. 1320-1327
発行日: 2006/12/31
公開日: 2009/03/27
ジャーナル フリーInterval arithmetic is applied to analysis of a nonlinear system, in particular, to computation of an output admissible set, which plays an important role in constrained control. A discrete-time system whose dynamics is expressed in polynomials is considered first. When the initial state of such a system is known to belong to some multi-dimensional interval, interval arithmetic enables us to compute an interval that includes all possible states after some time steps. This fact together with polynomial optimization is used to compute an output admissible set. Numerical examples show efficacy of the proposed approach. Extensions are considered to non-polynomial systems and continuous-time systems.抄録全体を表示PDF形式でダウンロード (2206K) -
非最小位相特性による最適制御性能の劣化について加嶋 健司2006 年 42 巻 12 号 p. 1328-1335
発行日: 2006/12/31
公開日: 2009/03/27
ジャーナル フリーIn this paper, by applying some results on infinite-dimensional control theory, a closed formula for the H2 performance limit is derived for systems that can be expressed as a series connection of a rational transfer matrix and a scalar inner function. This formulation is capable of dealing with various control problems, including H2 control of a class of infinite-dimensional systems. The resulting formula, given as a functional of the inner function, helps us to understand how achievable H2 performance deteriorates due to non-minimum phase properties of the plants. This paper also aims to clarify the performance limit via state-space representation, in order to handle multiple-input-multiple-output systems in a transparent way.抄録全体を表示PDF形式でダウンロード (1719K) -
木村 元2006 年 42 巻 12 号 p. 1336-1343
発行日: 2006/12/31
公開日: 2009/03/27
ジャーナル フリーIn real-robot applications, learning controllers are often required to obtain control rules over high-dimensional continuous state-action space. Random tile-coding is a promising method to deal with high-dimensional state space for representing the state value function. However, there is no standard reinforcement learning scheme to deal with action selection in high-dimensional action space, especially the probability of action variables are mutually dependent. This paper introduces a new action selection scheme using random tile-coding and Gibbs sampling, and shows the Q-learning algorithm applying the proposed scheme. We demonstrate it through a Rod in maze problem and a redundant arm reaching task.抄録全体を表示PDF形式でダウンロード (3026K) -
SLIPの提案と評価土谷 千加夫, 塩川 裕介, 池田 心, 佐久間 淳, 小野 功, 小林 重信2006 年 42 巻 12 号 p. 1344-1352
発行日: 2006/12/31
公開日: 2009/03/27
ジャーナル フリーReinforcement learning is a useful tool for complex control problems that cannot be modeled mathematically nor solved theoretically. However, a traditional value function approach such as Q-learning includes the difficulty of combinatorial explosion. Direct policy search, (DPS) is an alternative approach that represents a policy using some model and searches a parameter space directly for an optimum by optimization techniques such as genetic algorithms (GA). Instance-based policy (IBP) is a policy representation model of DPS. IBP represents a policy using a set of instances that are pairs of state and action. This paper presents a hybrid GA to optimize efficiently a set of instances with continuous state and continuous action, given an episodic task. The hybrid GA is composed of a combinatorial GA with BDX (Binomial Distribution Crossover) and a real-coded GA with INDX (Instance-wise Normal Distribution Crossover). The proposed method named SLIP (Sophisticated Learner for Instance-based Policy) was applied to a cat twist problem and a parallel-type double inverted pendulum problem.The results of experiments show the effectiveness and usefulness of SLIP.抄録全体を表示PDF形式でダウンロード (2457K) -
田中 良幸, 神田 龍馬, 武田 雄策, 山田 直樹, 福庭 一志, 正守 一郎, 辻 敏夫2006 年 42 巻 12 号 p. 1353-1359
発行日: 2006/12/31
公開日: 2009/03/27
ジャーナル フリーThe present paper investigates human mechanical impedance during the operation of a virtual steering system according to dual-arm configurations. The developed steering system using an impedance-controlled robot can accurately estimate mechanical impedance properties around the rotational axis. Experimental results demonstrate that a human changes his/her impedance properties in both the grip position on a steering wheel and the magnitude of steering torque. Overall stability of a human-steering system is then discussed by means of the measured human impedance parameters.抄録全体を表示PDF形式でダウンロード (2684K)
- |<
- <
- 1
- >
- >|