Cooperative manipulation of a floating object by some space robots: application of a tracking control method using the transpose of the generalized Jacobian matrix

In future space missions, it is considered that many tasks will be achieved by cooperative motions of space robots. For free-floating space robots with manipulators, we have proposed a digital tracking control method using the transpose of the generalized Jacobian matrix (GJM). In this paper, the tracking control method using the transpose of the GJM is applied to cooperative manipulations of a floating object by space robots. Simulation results show the effectiveness of the control method. This work was presented in part at the 12th International Symposium on Artificial Life and Robotics, Oita, Japan, January 25–27, 2007     

Digital control of space robot manipulators with velocity type joint controller using transpose of generalized Jacobian matrix

For free floating space robots having manipulators, we have proposed a discrete-time tracking control method using the transpose of Generalized Jacobian Matrix (GJM). Control inputs of the control method are joint torques of the manipulator. In this paper, the control method is augmented for angular velocity inputs of the joints. Computer simulations have shown the effectiveness of the augmented method. This work was presented in part and awarded as Best Paper Award at the 13th International Symposium on Artificial Life and Robotics, Oita, Japan, January 31–February 2, 2008     

A Secure Agent-based Framework for Internet Trading in Mobile Computing Environments

Most of the current Internet trading frameworks, in particular their negotiation and payment phases, are intended for customers frequently connected to the Internet during an entire transaction. This requirement cannot be easily met in the high communication cost and/or low bandwidth settings, typically found in mobile computing environments. Based on the software agent paradigm, a new secure agent-based framework for Internet trading in mobile computing environments is proposed in this paper. The framework is composed of two new protocols. One is the agent-based auction-like negotiation protocol, another is the agent-based payment protocol. Both of them are dedicated to solve the trade problems of Internet trading in mobile computing environments and ensured to be safe by cryptographic technologies. The combination of the two secure protocols constitutes an integrative solution for Internet trading in mobile computing environments.     

Performance comparison of deep learning architectures for surgical instrument image removal in gastrointestinal endoscopic imaging

Artificial Life and Robotics - Endoscopic images typically contain several artifacts. The artifacts significantly impact image analysis result in computer-aided diagnosis. Convolutional neural...     

Numerical Simulation for Direct Tunneling Current in Poly-Si-Gate MOS Capacitors

We have numerically simulated gate tunneling current in MOS capacitors. Price has demonstrated that the Gamow formulation can be applied to analysis of the escape of electrons from channel into gate in MOSFETs [P.J. Price, Appl. Phys. Lett., 82, 2080 (2003)]. We have integrated the Gamow method into a Schrödinger-Poisson solver for metal-gate and poly-Si-gate n-type MOS capacitors. The numerical results of the tunneling current are then compared with experimental results.     

Effect of magnetic iron impurity on the superconducting properties of an amorphous Nb50Zr35Si15 alloy

The effect of magnetic iron impurity on the superconducting properties of amorphous Nb₅₀Zr_35−x Si₁₅Fe _x (x⩽4 at %) alloys was examined. Doping with an iron impurity resulted in a linear depression ofT _c andH _c2(T) and a decrease in andρ _n after reaching a maximum value at 0.5 to 1.0 at % iron. The observed decrease was about 35% forT _c, 85% forH _c2 at 2.0 K, 16% for and 21% forρ _n. Although the decrease in occurs through the decrease inρ _n as expected from the GLAG theory, the depression inT _c caused by magnetic impurity could not be explained in terms of the GLAG theory which is applicable to Nb-Zr-Si amorphous alloys without magnetic impurity, but was interpreted as arising from the pair-breaking effect in the superconducting nature due to magnetic scattering. However, the pair-breaking effect was found to be smaller by about one-tenth for the present amorphous superconductors than for crystalline superconductors, indicating the high stability of the superconductivity of the Nb-Zr-Si-Fe alloys against the magnetic scattering arising from the magnetic impurity. The reduced magnetic field at which the reduced fluxoid pinning force exhibits a maximum value increased with iron concentration, indicative of an enhancement of fluxoid pinning force. The enhancement in fluxoid pinning force was interpreted as arising from the increase in compositional, electronic and/or magnetic fluctuations by the dope of iron impurity.     

Vertex angle and crossing angle resolution of leveled tree drawings

We study the problem of computing leveled tree drawings, i.e., straight-line drawings of trees where the vertices have distinct preassigned y-coordinates. Our optimization goal is to maximize the crossing angle resolution (i.e., the minimum angle formed by any two crossing edges) and/or the vertex angle resolution (i.e., the minimum angle formed by two edges incident to the same vertex) of the drawing. We provide tight and almost tight worst case bounds for the crossing angle resolution and for the total angle resolution (i.e., the minimum of crossing and vertex angle resolution), respectively.     

Adaptive control of underwater vehicle-manipulator systems using radial basis function networks

This paper deals with a control scheme for underwater vehicle-manipulator systems with the dynamics of thrusters in the presence of uncertainties in system parameters. We have developed two controllers that overcome thruster nonlinearities, which cause an uncontrollable system: one is a regressor-based adaptive controller and the other is a robust controller. However, the structure of the adaptive controller is very complex due to the feedforward terms including the regressors of dynamic system models, and the error feedback gains of the robust controller with a good control performance are excessively high due to the lack of feedforward terms. In this paper we develop an adaptive controller that uses radial basis function networks instead of the feedforward terms. The replacement leads to a moderately high gain controller whose structure is simpler than that of the regressor-based adaptive controller.