A moving switching sequence approach for nonlinear model predictive control of switched systems with state‐dependent switches and state jumps

In this paper, we propose an approach for real‐time implementation of nonlinear model predictive control (NMPC) for switched systems with state‐dependent switches called the moving switching sequence approach. In this approach, the switching sequence on the horizon moves to the present time at each time as well as the optimal state trajectory and the optimal control input on the horizon. We assume that the switching sequence is basically invariant until the first predicted switching time reaches the current time or a new switch enters the horizon. This assumption is reasonable in NMPC for systems with state‐dependent switches and reduces computational cost significantly compared with the direct optimization of the switching sequence all over the horizon. We update the switching sequence by checking whether an additional switch occurs or not at the last interval of the present switching sequence and whether the actual switch occurs or not between the current time and the next sampling time. We propose an algorithm consisting of two parts: (1) the local optimization of the control input and switching instants by solving the two‐point boundary‐value problem for the whole horizon under a given switching sequence and (2) the detection of an additional switch and the reconstruction of the solution taking into account the additional switch. We demonstrate the effectiveness of the proposed method through numerical simulations of a compass‐like biped walking robot, which contains state‐dependent switches and state jumps.     

Nonlinear Bayesian filtering via holonomic gradient method with quasi moment generating function

A symbolic-numeric method is proposed for addressing the Bayesian filtering problems of a class of discrete-time nonlinear stochastic systems. We first approximate the posterior probability density function to be Gaussian. The update law of the mean and variance is formulated as the evaluation of several integrals depending on certain parameters. Unlike existing methods, such as the extended Kalman filter (EKF), unscented Kalman filter (UKF), and particle filter (PF), this formulation considers the nonlinearity of system dynamics exactly. To evaluate the integrals efficiently, we introduce an integral transform motivated by the moment generating function (MGF), which we call a quasi MGF. Furthermore, the quasi MGF is compatible with the Fourier transform of differential operators. We utilize this compatibility to decrease the number of computations of Gröbner bases in the noncommutative rings of differential operators, which reduces the offline computational time. A numerical example is presented to show the efficiency of the proposed method compared to that of other existing methods such as the EKF, UKF, and PF.     

An estimation method of end-point impedance based on bilateral control system

ABSTRACT

This paper aims at developing an estimation method of end-point impedance. Human operators are constantly changing their end-point impedance for adapting to the surrounded environment and executing some complicated tasks, and it is highly meaningful to investigate these variations for the further understanding of human motion. Most of the conventional researches, however, have considered non-contact-point impedance or tasks that is only holding the vibrated sticks due to the experimental constraints. This paper proposes the estimation method of end-point impedance by using bilateral control system. The extra signal is added to the force controller for the impedance estimation. In addition, the effect of the bilateral controller is estimated and removed from the impedance estimation process for securing the applicability of moving tasks. The proposed method was validated through simulations and an experiment. The experimental result showed that the end-point stiffness can be estimated properly even if the operator robot was moving and changed its end-point impedance.     

Development of a low-energy and high-current pulsed neutral beam injector with a washer-gun plasma source for high-beta plasma experiments

We have developed a novel and economical neutral-beam injection system by employing a washer-gun plasma source. It provides a low-cost and maintenance-free ion beam, thus eliminating the need for the filaments and water-cooling systems employed conventionally. In our primary experiments, the washer gun produced a source plasma with an electron temperature of approximately 5 eV and an electron density of 5 × 10(17) m(-3), i.e., conditions suitable for ion-beam extraction. The dependence of the extracted beam current on the acceleration voltage is consistent with space-charge current limitation, because the observed current density is almost proportional to the 3∕2 power of the acceleration voltage below approximately 8 kV. By optimizing plasma formation, we successfully achieved beam extraction of up to 40 A at 15 kV and a pulse length in excess of 0.25 ms. Its low-voltage and high-current pulsed-beam properties enable us to apply this high-power neutral beam injection into a high-beta compact torus plasma characterized by a low magnetic field.     

Influence of oxide grain morphology on formation of the CuO scale during oxidation of copper at 600-1000 °C

To explain the very slow growth rate of the CuO scale during copper oxidation at high temperatures above 850 °C as reported in the literature, the influence of the morphology of the CuO grains on scale growth was investigated by oxidizing copper with different purities in 0.1 MPa oxygen atmosphere at 600-1000 °C. Oxidation of 99.99% copper shows that the growth of the CuO scale depends on the morphology of the CuO grains. The very thin CuO scale observed at 1000 °C is related to flat CuO grains resulted from fast lateral growth of the CuO grains, whereas the thicker CuO scale at lower temperatures results from the fine CuO grains. This dependence of scale growth on oxide grain morphology was confirmed by oxidizing 99.9999% and floating zone refined (>99.9999%) copper at 800 °C. It reveals that grain boundary diffusion can favour the growth of the CuO scale, and its contribution is related to the morphology of the CuO grains depending on the purity of copper specimens.     

New method for hydroxyapatite coating of titanium by the hydrothermal hot isostatic pressing technique

A new hydrothermal method is proposed, which enables us to prepare thin hydroxyapatite (HA) ceramic coatings on Ti substrates with a curved surface at low temperatures. The method uses double layered capsules in order to produce a suitable hydrothermal condition; the inner capsule encapsulates the coating materials and a Ti substrate, and the outer capsule is subjected to isostatic pressing under the hydrothermal condition. In this study, it is demonstrated that a pure HA ceramic layer with the thickness of 50 μm could be coated to a Ti cylindrical rod at the low temperature as low as 135 °C under the confining pressure of 40 MPa. The HA coating layer had a porous microstructure with the relative density of approximately 60%. Pull-out tests were conducted to obtain an estimate for the adhesion properties of the HA coating prepared by the double capsule method. The shear strength obtained from the pull-out tests was in the range of 4.0–5.5 MPa. It was also shown that the crack propagation occurred within the HA coating layer, not along the HA/Ti interface in the pull-out tests. This observation suggests that the fracture property of the HA/Ti interface was close to or higher than that of the HA ceramics only. It is expected that the low temperature double capsule method may provide a useful method for producing bioactive HA ceramic coatings on curved prostheses surfaces.     

A study of the loss characteristics of the amorphous magnetic material under nonsinusoidal wave magnetic flux

This paper presents a study on iron loss characteristics of an amorphous magnetic material under nonsinusoidal wave magnetic flux. The iron loss characteristics of the amorphous magnetic material have been compared to those of grain‐oriented magnetic steel sheets. Measurements of the iron loss have been carried out by means of an SST. It was found that the ratio of the iron loss increase of the amorphous magnetic material under the excitation by distorted waveforms and symmetric PWM waveforms is lower than that of grain‐oriented magnetic steel sheets. On the other hand, the ratio of the iron loss increase of the amorphous magnetic material under the excitation by asymmetric PWM waveforms is higher than that of grain‐oriented magnetic steel sheets. © 2004 Wiley Periodicals, Inc. Electr Eng Jpn, 150(3): 11–18, 2005; Published online in Wiley InterScience ( www.interscience.wiley.com ). DOI 10.1002/eej.20004     

Synthesis of nanotube from a layered H2Ti4O9 · H2O in a hydrothermal treatment using various titania sources

In this paper, the synthesis of tubular titania was carried out through a soft chemical hydrothermal reaction of TiO₂ powders in NaOH or KOH aqueous solution systems. It was found that nanotubular products prepared in our studies were identified as H₂Ti₄O₉·H₂O by X-ray diffraction analysis with their morphology and crystallinity being dependent on synthetic conditions, i.e. reaction time and temperature of the hydrothermal process. The photocatalytic activity of nanotubular H₂Ti₄O₉·H₂O was evaluated in a decomposition test of HCHO at 298 K in an aqueous system using radiation with a mercury lamp. The morphology and yield of these nanotubular products were found to be dependent on the hydrothermal synthetic conditions.     

A new characterization method of the microstructure using the macroscopic composition gradient in alloys

A new experimental method to determine the phase boundary and phase equilibria is accomplished by using the transmission electron microscopic observation of alloys having the macroscopic composition gradient. The various phase boundaries,i.e., the coherent binodal and spinodal lines and incoherent binodal line, are distinctly determined for the Cu-Ti alloy system. Furthermore, the equilibrium compositions at the interface of precipitate/matrix are experimentally obtained for various particle sizes, and therefore, the Gibbs-Thomson relation is verified. It is expected that the composition gradient method proposed in the present study will become an important experimental method for microstructural characterization.     

Proposal of voltage transient sag compensator with controlled gradational voltage

This paper presents a new concept for a voltage transient sag compensator and the experimental result of its 400‐V‐class compensator. This compensator is composed of the series connection of some inverter units with gradational output voltages. Because each output voltage is different by 2ⁿ times, an approximate sinusoidal voltage is generated by controlling operation of each inverter and compensates voltage sag of the power line. The compensator can be directly installed in a power line without an insertion transformer and a large filter, and thus it is expected to be a compact and economical system. © 2005 Wiley Periodicals, Inc. Electr Eng Jpn, 154(3): 65–72, 2006; Published online in Wiley InterScience ( www.interscience.wiley.com ). DOI 10.1002/eej.20289