Ryuta Ozawa

Adaptive neural network control of tendon-driven mechanisms with elastic tendons

We propose an adaptive control and an adaptive neural network control (composed of two RBF neural components and one adaptive component) for tendon-driven robotic mechanisms with elastic tendons. These controllers can be applied to serial or parallel tendon-driven manipulators having linear or non-linear elastic tendons. We begin by proving the stability of the adaptive control system for our mechanism, and then we prove the stability of the adaptive neural network system and report on the results of numerical simulations and experimental results performed using a 2-DOF tendon-driven mechanism having six elastic tendons.

Dynamic visual servoing with image moments for a quadrotor using a virtual spring approach

This paper presents an image-based visual servoing for controlling the position and orientation of a quadrotor using a fixed downward camera observing landmarks on the level ground. In the proposed method, the negative feedback to the image moments is used to control the vertical motion and rotation around the roll axis. On the other hand, the negative feedback cannot be used to control the horizontal motion due to under-actuation of a quadrotor. Thus, a novel control method is introduced to control the horizontal motion. Simulations are presented to validate the proposed method.

Design and control of underactuated tendon-driven mechanisms

Many robotic hands or prosthetic hands have been developed in the last several decades, and many use tendon-driven mechanisms for their transmissions. Robotic hands are now built with underactuated mechanisms, which have fewer actuators than degrees of freedom, to reduce mechanical complexity or to realize a biomimetic motion such as flexion of an index finger. The design is heuristic and it is useful to develop design methods for the underactuated mechanisms. This paper classifies mechanisms driven by tendons into three classes, and proposes a design method for them. The two classes are related to underactuated tendon-driven mechanisms, and these have been used without distinction so far. An index finger robot, which has four active tendons and two passive tendons, is developed and controlled with the proposed method.

Analysis, Classification, and Design of Tendon-Driven Mechanisms

This paper analyzes tendon-driven mechanisms (TDMs) with active and passive tendons and proposes a method for designing TDMs. First, we group TDMs into six classes according to their controllability and the number of driving degrees of freedom. In this classification system, the conventional underactuated mechanisms are grouped into three classes, two of which have often previously been grouped together although they have different manipulation abilities. Next, we analyze bias forces to separate and decouple a given TDM into several smaller TDMs. Finally, we propose a design method for combining smaller TDMs into an appropriate TDM. Using this method, we can easily determine an appropriate tendon transmission that meets the requirements for the number of tendons, the hardware of the tendon routing, and arbitrary joint constraint that has useful applications in prosthetic and biomimetic hands. Numerical examples show that the proposed method guarantees a nonsingular series actuation transmission and that the designed underactuated TDMs could achieve arbitrary stiffness independent of the actuation effort compared with the conventional underactuated TDM with torsion springs.

Two-Dimensional Stable Blind Grasping under the Gravity Effect

This paper is aiming at showing a sensory-motor coordination control scheme that realizes stable pinching of rigid objects with parallel or non-parallel flat surfaces movable in 2-dimensional vertical plane by a pair of robot fingers with hemispherical ends. The proposed control signal is composed of gravity compensation for fingers, damping shaping, exertion of forces to the object from opposite directions, generation of moments for rotational moment balancing, and regressors for estimating unknown steady-state terms, all of which neither need the knowledge of object parameters nor use any object sensing data. In other words, stable grasping can be realized by using only finger joint sensing in a blind manner without using force sensors and tactile sensing. Stability of pinching motion with convergence to the state of force/torque balance is shown through computer simulations and is also proved theoretically.

A new impedance control concept for elastic joint robots

The purpose of this paper is to propose a new impedance control concept for elastic joint robots with programmable passive impedance devices in the transmission. The concept allows us to use the same index both for free motion and for contact task. We apply it to a one-DOF elastic joint robot and derive an adjustment law for the robot. The numerical simulations show that implementation of the concept can be realized without any dynamic models.

Dynamic visual servoing with image moments for an unmanned aerial vehicle using a virtual spring approach

Abstract This paper presents an image-based visual servoing for controlling the position and orientation of an unmanned aerial vehicle (UAV) using a fixed downward camera observing landmarks on the level ground. In the proposed method, the visual servoing of the image moments is used to control the vertical motion and rotation around the roll axis. In contrast, an undesired positive feedback arises in visual errors because of the under-actuation of the UAV and this positive feedback makes it difficult to apply the visual servoing to the horizontal motion. Thus, a novel control method using the virtual spring is introduced to control the horizontal motion. The stability of the system is proved based on Lyapunov’s direct method. Simulations are presented to validate the proposed method.

Design and control of a three-fingered tendon-driven robotic hand with active and passive tendons

This paper presents a design of a three-fingered robotic hand driven by active and passive tendons and proposes control methods for this hand. The tendon-driven robotic hand consists of the thumb, the index and the middle fingers. The robotic thumb can move all the joints independently. In contrast, the index and the middle robotic fingers are under-actuated using the combination of active and passive tendons, and move the terminal two joints synchronously, which is one of the important features of the human digits. We present passivity-based impedance and force controllers for tendon-driven robotic fingers and discuss how to combine them for fast and secure grasps. We experimentally validate that the robotic hand moves fast and manipulates an object and demonstrate that the robotic hand grasps objects in diverse ways.

Stable grasping and relative angle control of an object by dual finger robots without object sensing

This paper proposes stable grasping and angle control methods of an object with parallel surfaces by a pair of finger robots in a horizontal plane. The method does not require any object sensing for stable grasping but only measurement of the object angle for absolute orientation angle control. Relative orientation angle control of the grasped object can be attained even if no object sensing is used.

Development of an ankle-foot orthosis with a pneumatic passive element

This paper presents a new ankle-foot orthosis (AFO) for hemiplegic patients. The proposed AFO has a pneumatic passive element with a rotation axis at the ankle joint. The ankle-joint constraint force can be adjusted according to the subjects gait phase. We confirmed that adjusting of the pneumatic pressure in the proposed pneumatic passive element enables the device to duplicate the performance of many conventional AFOs