A robust control approach is described for controlling the position and orientation of a six-DOF parallel link manipulator, in which one prismatic link consists of a rubber muscle actuator, so that the controlled object contains several nonlinear factors. The position of the link is also assumed to be measured by using a linear potentiometer. As a result, the associated measurement system contains some measurement noises. The control system is construted using a stochastic fuzzy controller with an integlator so as to remove a tracking error. To design the feedback gains of the stochastic fuzzy controller, a mmn-max control strategy is adopted to the systems, with unknown physical parameters. The effectiveness of the proposed method is illustrated by using some actual experiments.
This paper proposes a learning method which employs an assistance for simplifying a motion to learn and learns it from easy to difficult gradually. Using Genetic Algorithm, it searches for the parameters of a controller appropriate for controlling the motion to learn with gradually increasing difficulty, i.e., with gradually decreasing degree of assistance. We show that this gradual search enables Genetic Algorithm to evolve a population of controllers efficiently by giving two examples: stable riding of a bicycle and stable control of a double inverted pendulum. A bicycle is much easier to control when it is running at a certain velocity. An initial velocity is given as assistance. The learning method searches for a controller gradually decreasing the initial velocity. Similarly a double inverted pendulum is much easier to control when an upward force supports the distal end of the pendulum. The learning method searches for a controller gradually decreasing the upward force. The reduction rate of assistance is adjustable in accordance with the adaptability of a population to the reduction.
We discuss some suitable stability margin concepts for walking robots which are expected to walk around on irregular terrains. Several stability margin ideas have been proposed before, but those can be roughly divided into three kinds.We came to the conclusion that a stability margin concept based on energy considerations is the most practical to adopt and that its validity could be recognized through simple experiments. But the existing energy stability margin is inadequate for practical use because this margin varies with the weight of the robot which can't affect its resistance to tumbling caused by slipping from steps or the sudden stop. Therefore, we'll propose an improved one here, called the normalized energy stability margin, the NE stability margin for short. Considering this margin, we'll introduce a novel tool called the SNE contour which connects the points on the slope which have the same NE stability margin, so as to derive a stable gait for walking robots. In the final part of this paper, we focus on the features of the SNE contour which shed new light on the solution for the stable posture.
This paper develops a sensor-based navigation system using a target direction sensor for mobile robots. The importance of target direction in navigation problem for mobile robots is emphasized. Then, The convergence property for sensor-based navigation problem using the target direction is discussed. An actual sensor system using two CdS sensors is proposed. An navigation algorithm using the sensor system is also presented. The algorithm uses only two kinds of information by two sensor systems, which are target direction and existence of obstacles near the mobile robot, unlike typical sensor based navigation algorithm which uses the information of the position of the mobile robot. Some experiments by a developed mobile robot using the sensor-based navigation algorithm show an effectiveness of the navigation system and the algorithm.
Much attention has been paid recently to Augmented Reality (AR) which supports human's daily life by blending multimodal information into real world, typically on the objects being considered. Most of existing AR systems provide such information to human by using ordinary displays and see-through head mounted displays. However, monitoring and understanding of human actions is necessary for the system to achieve flexible and gentle interactions. The paper presents an attempt to make such interface using an extended digital desk for guiding and teaching robot system. A prototype system consists of a projector subsystem for information display and a realtime tracking vision subsystem for recognizing human's action. We implemented coordination of two levels of interactions using a virtual teaching panel and projected real image of the task environment on a desk for intuitive teaching. Experiments of teaching robot tasks demonstrate usefulness of the proposed system.
A multiple-sensor-enhanced robot collaboration system has been developed for manufacturing applications. The system consists of basic robot modules, such as for locating and tracking. These modules have laser range finders and perform hybrid manufacturing tasks by working together as an advanced robotic cell. A hybrid task with uncertainties can be described by the task velocity and task error derived from the sensory and motion information on the robot modules. The modules send their task error and waiting time to each other via a collaboration network. They also modify their given motion sets in real time based on fuzzy rules in order to improve the accuracy and efficiency of the system. An experimental study of multi-station welding showed that this collaboration results in a satisfactory trade-off between accuracy and efficiency.
We developed a high-performance actuator using particle-type ER fluid. This ER actuator have high torque/inertia ratio, and can respond quickly. In this paper, the force control properties of this actuator were examined experimentally. Firstly, we developed the model of ER actuator based on experiments. Then the feedback control system is designed by H∞ control theory. Lastly, the proposed control system is examined by experiments.
Small and precise actuators for multi-degrees of freedom (DOF) motion is needed to construct a dexterous robot arms and manipulators. Ultrasonic motors are suitable for the multi-DOF actuation of the robot arms and manipulators because of its characteristics such as its output power per unit volume, high stationary limiting torque, high maximum operating torque, simple design, silence, and high controllability. In this study we develop a three-DOF ultrasonic motor. It is found that the three-DOF ultrasonic motor can be constructed when natural frequencies of two second bending modes and a first longitudinal mode of a bar-shaped stator correspond. A spherical rotor in contact with the stator head rotates around three perpendicular axes. Geometry and structure of the ultrasonic motor in detail is determined using finite element analysis. Measured natural frequencies and natural modes of bending and longitudinal vibrations of the stator agree well with calculated ones. The rotational speed of the spherical rotor around three perpendicular axes are measured.
This paper proposes a planning method for time-optimal trajectories of mobile robots. A dynamical model for two independently driven wheels type mobile robots is derived. Using an idea of path parameter, optimal trajectory for specified path of mobile robot is easily obtained considering dynamical constraints such as driving torques and velocities. An efficient collision-free near-time-optimal trajectory planning method is proposed using a local optimization method and a global search method of initial paths. The B-spline parameter optimization method is used as a local optimization. Some numerical examples show an effectiveness of the proposed planning method.
Manipulators have similar structure with human upper extremity, and there are many possibilities to utilize them for various works instead of humans. Hence, human arm movements have been widely studies, which analyze sophis-ticated capability of human arm control system and attempt to improve ability of manipulators by transferring the results of such analysis. It is reported that human arm trajectories have particular properties such that most of its path lies along a rather straight or simple curved line, and a tangential velocity of hand is bell shaped. In this paper, human arm trajectories are analyzed in Cartesian space and joint angle space, and performance criterion is proposed to formulate trajectories for demonstrating human arm movements. Moreover, difficulty of solving nonlinear optimization problem which disturbs manipulators from application of trajectory planning is overcome by developing an algorithm of near-optimal trajectory formation based on approximated differential equations. Besides, it is confirmed that optimal trajectory with high accuracy is stably formulated by utilizing the algorithm Formulated trajectories according to a proposed trajectory planning strategy are compared with observed trajectories obtained by experiment which is conducted under unrestricted conditions. The results show that the formulated trajectories well demonstrate human arm trajectories.
In the present study we attempt to induce a quadruped robot to walk dynamically on irregular terrain and run on flat terrain by using a neural system model. For dynamic walking on irregular terrain, we employ a control system involving a neural oscillator network, a stretch reflex and a flexor reflex. Stable dynamic walking when obstructions to swinging legs are present is made possible by the flexor reflex and the crossed stretch reflex. A modification of the single driving input to the neural oscillator network makes it possible for the robot to walk up a step. For running on flat terrain, we combine a spring mechanism and the neural oscillator network. It became clear in this study that the matching of two oscillations by springs and the neural oscillator network is important in enabling the robot to hopping. The present study also shows that entrainment between neural oscillators causes the running gait to change from hopping to bouncing. This finding renders running fairly easy to attain in a bounce gait. It must be noticed that the flexible and robust dynamic walking on irregular terrain and the transition of the running gait are realized by the modification of a few parameters in the neural oscillator network. This finding, obtained through experiments using the quadruped called Patrush, shows the potential ability of a neural oscillator network to facilitate adaptation in dynamic walking and running.
The importance of force and tactile display in virtual reality technology has been recognized recently and numerous researches have been done including development of various force and/or tactile display devices. However, many of them are always in touch with fingers or hands of operators and it is impossible for an operator to feel difference between contact and non-contact state with a virtual object directly from sensory channel of touch in his/her finger or hand. This paper presents a new haptic display device we have developed which provides both touch and force feeling to the operator's fingers. This device tracks the operator's finger with no contact when the finger is not in contact with any virtual object. It touches and displays force to the finger only when some virtual object is in touch with the finger. A control algorithm for the system to display dynamic virtual objects is given. Preliminary experimental results are also presented.
This paper discusses the dynamics computation of structure-varying kinematic chains which imply mechanical link systems whose structure may change from open kinematic chain to closed one and vice versa. The proposed algorithm can handle and compute dynamics and motions of any rigid link systems in a seamless manner without switching among algorithms. The computation is developed on the foundation of the dynamics computation algorithms developed in robotics, which is inherently superior in efficiency due to explicit use of the generalized coordinates to those used in the general purpose motion analysis software. The structure varying kinematic chains are commonly found in computing human and animal motions. The established computation will provide the general algorithmic foundation of the computation of motion and control of humanoid robots and CG human figures.