Many researchers are working on nonholonomic systems modeled as driftless nonlinear state equations, such as vehicles and flying robots with zero initial angular momentum. Flying robots with nonzero initial angular momentum, however, are no longer modeled as driftless state equations and do not attract much attentions so far. Although they have drift terms, they are still tough systems for feedback control. In this paper, we will propose an attitude controller for flying robots with nonzero initial angular momentum. We will show that, by introducing time varying state coordinate transformation and feedback, the state equation of the flying robot can be transformed into a time invariant nonlinear state equation whose linear approximation is controllable. This state equation enables us to control the landing posture of the flying robot. An adaptive controller is also proposed to eliminate the posture error caused by parameter uncertainties.
Accurate oculomotor control is one of the essential pre-requisites of successful visuomotor coordination. Given the nonlinearities of the geometry of binocular vision as well as the possible nonlinearities of the oculomotor plant, it is desirable to accomplish accurate oculomotor control through learning approaches. In this paper, we investigate learning control for a biomimetic active vision system mounted on a humanoid robot. By combining the adaptive control strategy of feedback-error learning with a state-of-the-art statistical learning network, our robot system is able to acquire high performance visual stabilization reflexes after about 40 seconds of learning despite significant nonlinearities and processing delays in the system.
This article, as a fundamental theory for dealing with intelligence of robots, discusses a fuzzy logic satisfying Boolean properties. This article shows the following: General construction of Boolean logics, and significance of each property that should be satisfied so that a whole set of well-formed formulas may form a Boolean algebra; Outline how to process fuzzy inference through an intuitively-understandable example that implies why usual fuzzy logics are not suitable for “deep inference” and what effect Boolean properties have on inference processes; Definition of the extended fuzzy subsets modified from usual fuzzy subsets such that membership functions take binary sequences instead of truth values; Construction of a Boolean fuzzy logic based on extended fuzzy subsets that is one of fuzzy logics satisfying all of the Boolean properties; Process of updating inference when a new knowledge is supplemented; Techniques of calculating membership values in extended fuzzy subsets on condition that some truth values are given; Boolean properties concerning extended fuzzy subsets; How to advance fuzzy inference on the Boolean fuzzy logic.
Development of new type of mobility is required for space robotics projects such as The MUSES-C aiming at small asteroid exploration. Since the environment of asteroids is of vacuum and microgravity, floatation is a possible choice for enhancing a new mobility of the rover. In this paper, we propose the use of electro-magnetic levitation in order to integrate a mobility into the microgravity rover. The rover has a spherical shape and a smaller spherical shell inside. Four electro-magnets are symmetrically located between the outer sphere surface and the inner sphere shell with one end of each directed to the center of the shell. With electro-magnetic force of the magnets, a sphere iron ball inside the shell is controlled and levitated. When the rover lifts the ball inside with the electro-magnetic force, the rover is in return pressed down the ground by the reaction force, due to which the rover system not only gains upward momentum for floatation, but also obtains friction that enables its rolling on the ground. The prototype microgravity rover was developed and provided experimental results that indicates effectiveness of the proposed mobility.
During teleoperation, when the operator's attention is concentrated in controlling the end-effector motion, it is very difficult to recognize a possible link collision. Such collision may damage the environment or the link seriously, because usually the link has no sensors. In this paper, we propose a so-called“virtual radar”which displays the collision information considering the whole manipulator at the end-effector level. The computer checks for unreachable areas due to the presence of obstacles and displays the appropriate information mapped in task-space coordinates using a 3D graphical image. This enhances the reliability of the system. However, in the case of a 6 D.O.F manipulator, collision possibility space must be a 6 dimensional one and hence it is difficult to visualize and understand. To tackle this problem, a deformed sphere and patterns on its surface are employed. We have applied this method to a 6 D.O.F. manipulator and have verified the effectiveness of the proposed approach by experiments with a real telerobotic system.
Manipulators with free joints are second-order nonholonomic systems whose dynamical constraints are noninte-grable. Such systems are known to be difficult to control and are underactuated systems which have possibility to steer many joints by only one motor. Previously proposed methods to control free-joint manipulators have been based on an assumption of perfectly frictionless free joints. In this paper, 2R and 3R free joint manipulators with only one actuated joints are to be studied with consideration of friction at free joints. Averaging analysis clarifies the frictionless free-joint manipulators are Hamiltonian systems with a conservation of an energy-like quantity. Various models of friction at the free joint are considered for 2R free-joint manipulators and the averaging analysis and simulations reveal that the systems with friction show dissipative behaviors. For 3R free-joint manipulators, the Poincaré map of their frictionless behaviors in response to periodic inputs show torus-like invariant manifolds in 4D phase space. From experiments and simulations, 3R free joint manipulators with friction show the behaviors converging to an equilibrium point. Considering these dissipative behaviors with friction, a method using the energy-like quantity to stabilize to the equilibrium point is proposed for 3R free joint manipulators. Several control methods to position to any destination via stabilization to desired invariant manifold are proposed for 2R free-joint manipulators and simulations and experiments show their effectiveness.
We are trying to induce a quadruped robot to walk dynamically on irregular terrain by using a nervous system model. In this paper, we integrate several reflexes such as stretch reflex, vestibulospinal reflex, and extensor and flexor reflex into CPG (Central Pattern Generator) . We try to realize adaptive walking up and down a slope of 12 degrees, walking over an obstacle 3 [cm] in height, and walking on terrain undulation consisting of bumps 3 [cm] in height with fixed parameters of CPG and reflexes. The success in walking on such irregular terrain in spite of stumbling and landing on obstacles shows that the biologically inspired control proposed in this study has an ability of autonomous adaptation to unknown irregular terrain.
In our previous work, we have proposed a mixed force and motion commands based space robot teleoperation system and a compact 6-DOF haptic interface to achieve an effective manual teleoperation. Until now, the effectiveness of this system and the haptic interface have been confirmed by the experiments in our laboratory. The most important features of this teleoperation system are a robustness against modeling errors and an ability to realize an operator's force at a remote site. In this paper, elements of this system and the haptic interface have been employed to teleoperate the 6-DOF manipulator mounted on the real space robot system ETS-VII. Surface tracking and peg-in-hole tasks have been executed to confirm the effectiveness of our system and the haptic interface. The results show that our space robot teleoperation system including the haptic interface has been able to execute these tasks in real space without any problems.
In jumping motions of humans, swings of arms and counter movements of legs are essential. In this paper, the swing of arms is focused. A pendulum type jumping machine is considered to model the swing of arms, which consists of two articulated links and a joint actuator, and a jumping motion with the swing of the second link is performed. The possibility of a soft landing with the swing of the second link is also investigated. A control scheme for the soft landing of the pendulum-type jumping machine is proposed, and several experiments of posture control for the soft landing demonstrate the feasibility of the control scheme.
Considering that the robots work in real environment, it is very important to recognize its condition by itself. Therefore, we are developing a self-diagnosis system for the autonomous robots. Especially, in this paper, we develop the basic sensory system and self-diagnosis method of an autonomous mobile robot. At first, we discuss the definition and classification of system accident. We also propose actual algorithm and method to diagnosis internal status of the robot using sensory information. The developed sensory system for self-diagnosis consists of multiple sensors, which are equipped on input/output node of each component. We show some experimental result using real autonomous mobile robot and our proposed diagnosis algorithm.