This paper describes possible factors of spatial error distribution in a given indoor space based upon a local positioning system which utilizes spread spectrum ultrasonic waves generated by narrow directional ultrasonic transmitters. The positioning error has been interestingly discussed from the viewpoints of the dilution of precision (DOP) and the ranging error mainly due to the transmitter directional characteristics. The indoor 3-D position is calculated by means of an algorithm using distances from the transmitters located at the four corners of the ceiling to a mobile receiver. As the result of this experiment, it has been usefully found that the positioning error could be more largely influenced in the vertical direction DOP than the horizontal ones, and that the obtainable positioning error could be averaged to 3.43 [cm] rms for all 3-D measured points. These findings obtained in this work imply that the 3-D indoor positioning can be realized with an acceptable centimeter-order of precision. More importantly, the 3-D indoor positioning has to be prepared with the viewpoints of not only the ranging error, which could be yielded by the distance measurement, but also DOP, which is inevitably given by a geometrical arrangement of transmitter/receiver relation.

抄録全体を表示

Organic gas adsorption onto optical surfaces equipped in satellites is one of the causes of signal degradation occurred in an orbit. To estimate the transmittance degradation caused by gas adsorption onto optical glass surfaces, the transmittance measurements were carried out in UV-VISUAL (200-800nm) and IR (1.7-15μm) wavelength ranges.Five kinds of glasses, SiO₂, BK7, Al₂O₃, CaF₂, and ZnSe, which are typically used for satellite optics, were selected as glass samples and three kinds of organic gas, 2-propanol, ethyl-acetate, and dichloromethane, which are dominant components of organic solvent, were selected as adsorption gases. In infrared wavelength range, several absorption bands correspond to vibration and/or bending transitions of functional groups included in adsorption molecules were detected. The strengths of these absorption bands decreased with glass temperature depression and completely disappeared at 233K. In UV-VISUAL wavelength, there is no local absorption features but broad transmittance degradations were detected. The relative spectral degradation for dichloromethane is similar to that for 2-propanol, but not similar to the case for ethyl-acetate. The differences of transmittance degradations among three kinds of adsorption molecules can be explained by the bonding structure of molecules. Namely, ethyl acetate has a π-bond, which can absorb visual light with π-π transition, and 2-propanol and dichloromethane have no π-bond but σ-bonds, which can absorb UV light. From the comparison of measured spectral transmittance degradations with sensor output degradations, it is found that the signal degradations of launched sensors are more similar to the transmittance degradation due to the 2-propanol or dichloromethane adsorption than that with ethyl-acetate. Moreover, we estimate the growth rate of the adsorbed molecular film thickness using the degradation data of orbiting sensor, MODIS/Aqua. To carry out our estimation, we adopted the assumption that the signal degradation is caused by the transmittance decrease of the sensor optics due to the gas adsorption. Our estimation shows that the growth rate of adsorbed molecular film during the period from 30 to 50 months after the launch is significantly lower than the growth rate just after the launch.

抄録全体を表示

This paper deals with stabilization of continuous-time systems with ternary symbolic inputs. We consider to generate the control input based on PWM (pulse width modulation) whose duty ratio is given by a state feedback control law. The analysis and synthesis is done based on the discrete-time domain. Although the discretized system nonlinearly depends on duty ratio, a stabilizing state feedback gain and a corresponding invariant set can be obtained by using linear matrix inequalities. Moreover, we show that the origin can be exponentially stabilizable by using PWM-type control input for almost all the stabilizable continuous-time systems.

抄録全体を表示

This paper proposes a new synthesis of a gain scheduling controller. Using the dilation lemma, the synthesis condition is derived in terms of linear matrix inequalities with a line-search parameter, which can be solved numerically. Although previous design methods guarantee control performance for every parameter value, resultant gains tend to be too complicated to be implemented in a real system. In contrast, our method yields simple scheduling laws which are preferable for implementation. Derived scheduling laws are piecewise-linear on scheduling parameters and have fewer grid points than scheduling laws of a previous method. Since the synthesis is based on parameter-dependent Lyapunov functions, simple scheduling laws can be derived without deteriorating the performance of the closed loop system. A numerical example demonstrates the effectiveness of the proposed method.

抄録全体を表示

The Engineering Test Satellite VIII (ETS-VIII) spacecraft has been launched in Geo-stationary orbit in December 2006 by Japan Aerospace Exploration Agency (JAXA). Its main mission is a mobile communication of the next generation. For the purpose, it has two large deployable antenna reflectors and a pair of large solar paddles that rotate around the pitch axis. We plan to perform advanced on-orbit attitude control experiments at the end of its mission life. The experiment aims to apply advanced control theories in order to develop a baseline for future controller design technology for this class of satellite. This paper proposes 2-degree-of-freedom control based on robust direct velocity and displacement feed-back as a candidate. The spacecraft modeling and controller synthesis methods are discussed. Then, we show some numerical study results.

抄録全体を表示

In this paper a new design method for a gain scheduling controller is described which will be employed in an on-orbit control experiment using Engineering Test Satellite VIII. Using this method, a robust control system that is resilient to changes in the paddle rotation angle and higher order vibration modes is designed as this method can treat the satellite system as a linear parameter varying (LPV) system whose dynamics vary according to the paddle rotation angle. The derived gains can be easily scheduled since the scheduling law consists of small number of grid points and requires just linear interpolation calculation. Using this method, stability and performance are guaranteed at each parameter value (paddle angle in this case). Some simulation results are provided to show the effectiveness of our controller.

抄録全体を表示

This paper presents an application of single-input multi-output (SIMO) system identification in the time domain using accelerometer outputs that are typical of satellite on-board sensors.The algorithm is suitable for the on-orbit system identification of six degrees-of-freedom spacecraft motions using the responses to thruster impulse inputs. Our new method identifies the input and output matrices of a SIMO collocated system by applying a recursive least-squares iteration scheme to refine the matrices obtained from conventional ERA. In a numerical simulation of an actual spacecraft to verify the algorithm, a mathematical model of the Engineering Test Satellite-VI (ETS-VI) is excited by a body-mounted thruster, and the translational and rotational accelerometers are measured simultaneously. Our method successfully identifies all of the natural frequencies, damping ratios, and mode shape vectors except for pitch rotational motion that has low sensitivity, confirming its validity.

抄録全体を表示

Pneumatic actuators are suitable for humanoid robots interacting with persons in terms of size and safety since they have high flexibility and high power to weight ratio. However, it is known that it is difficult to control them stably by a feedback control because of their large dead time and high-order delay. Therefore, when we apply the feedback error learning to learn an inverse model of a system includes such actuators, the strong non-linearity makes the learning unstable. This paper focuses on an android which adopts pneumatic actuators with strong non-linearity and proposes a method to learn the inverse model of the android by compensating the characteristics of the actuators.

抄録全体を表示

Human can pinch, grasp and manipulate objects stably and dexterously. It is well known that to these skills a unique configuration in human hand, called “Fingers-Thumb Opposability”, contributes. This paper focuses on the configuration and function of a thumb, and treats a simultaneous force/position control of contact without any external sensing, called, “Proprioceptive Touching”, by means of a5 D. O. F. robotic thumb with soft and deformable hemispherical finger-tip under nonholonomic rolling constraints. Firstly, 3-Dimensional nonholonimic rolling constraints with deformation of the finger-tip coming from its softness are formulated, and an overall dynamics that contains these rolling constraints explicitly is derived. Next, “Proprioceptive Touching”controller is designed and the convergence of the closed loop dynamics to an equilibrium point is proved. Numerical simulations are performed to illustrate that our proposed controller makes it possible to realize blind touching no matter when the finger-end contacts the task plane.

抄録全体を表示

In this note, we consider the optimality of Fictitious Reference Iterative Tuning (F RIT), which is one of the methods for tuning the parameter of a controller using only one-shot experimental data and without using mathematical models. We give a generalized condition for the prefilter which is performed to the initial data to yield the desired parameter. We then show that this genealized condition contains not only the previous work by the author but also the result in Virtual Reference Feedback Tuning (VRFT) as its special cases.

抄録全体を表示