This paper describes composite sensor for industrial robot, which detects both distance and slip. The fringe of interference by two laser beam is made by the sensor on the surface of an object which robot will hold. When the sensor is moved pararell to the surface of an object, the intensity of reflected light from the position of fringe changes periodically. And distance is calculated from the period of signal, which is proportional to distance. When slip occurs after robot holds an object, signal changes peiodically again because an object moves to the sensor relatively, and slip velocity can be calculated. Shape, position and orientation of an object can also be detected by measuring distance to several points.
This paper describes a collision avoidance method for a mobile robot operated by history of human operators as well as by computerized fuzzy controller. In order to cope with changeable environment, a mobile robot has to provide higher planning which watch success, failure of plan and provide feedback of perception. In this paper, a mobile robot makes a judgment based on relative position and speed between a robot and obstacles, and it predicts next environment to some extent. Moreover a mobile robot tries to acquire a skill-based intelligent through layered neural network. Some results of computer simulation are shown to ascertain effectiveness of this two-stage avoidance control. Finally, we propose layered pass-planning by thinking level which considered robot dynamics.
The purpose of this paper is to build an autonomous distributed robot system composed of mobile robots of the same type with the mobility of micro-mice and with a simple communication apparatus. The feature of the proposed robot system is its behavior learning capability by a distributable genetic algorithm (dGA) . dGA differs from GA in the treatment of fitnesses of individuals and can be implemented on a group of mobile robots. The robot behaviors are modeled by simple finite automata and their transition functions are encoded into genes in each individual robot. The transition functions determine the performance of the robots and are subject to genetic operations according to their results with a generation in dGA as a trial in the learning process. A simulation with 100 individuals (4-state 2-input automata) verifies the validity of the proposed approach. The behavior learning is found to improve the performance of the robot system as a whole.
The approach presented in this paper is based on two ideas : (1) on the planning level an electrostatic potential field in the configuration space of the robot-manipulator is used. The generalized force curves are attractive to the goal point avoiding obstacles without local minima ; (2) on the control level the system follows these force curves using sliding mode control. The system is attracted to the swishing manifold using sliding mode and therefore the system is robust to disturbances and parameter variations.
Dynamic hybrid control of robot manipulators considering object dynamics are investigated. Manipulator and object are modeled using lumped parameters. First, controllability and observability of the system in a one-dimensional case are clarified. Second, the optimal servo controller which merges an observer is designed, and the advantage of the controller is shown comparing with a controller which is designed neglecting object dynamics. Finally, the theory is extended to hybrid control in a two-dimensional case, and it is shown that the proposed controller is effective for hybrid control as well as force control.
A system for displaying the operating feel while manipulating virtual objects which takes their dynamics into account is introduced. The display is realized using multiple link mechanisms. New concept of impedance display and two ways of its realization, “measuring force and displaying motion”and“measuring motion and displaying force”, are proposed. Calculation algorithms of driving input to the display device for the operation of both single-body and multi-body virtual objects are established. The validity of the proposed method is verified experimentally using a newly developed two-finger display device.
This paper discusses kinematics and control problems of a robot wrist mechanism with redundant muscles (actuators) . The redundancy enables us to reduce mechanical backlash, avoid near zero tendon forces, where the control accuracy goes down, and adjust the mechanical compliance. Because of the adjustability of the mechanical compliance, the wrist can adapt itself to changeable task environments easily and reduce vibrations by disturbances. On the other hand, high antagonistic forces and slacks cause the system to be unstable. The wrist has a novel mechanism to overcome some problems in a conventional tendon-controlled mechanism. That is, nonlinear spring tensioners (NST) are inserted in a drive path to give the wrist more ability of adjustment of the mechanical compliance than a conventional tendon-controlled mechanism. Since desired mechanical compliance is easily realized by setting up the displacement of springs, the burden of controllers becomes very little. Therefore the wrist can perform with smaller stiffness and faster response than a conventional tendon-controlled mechanism. In this paper, a control strategy for this mechanism is also discussed. Experimental results are given to show the efficiency of the tendon-controlled wrist mechanism with NST.
In this paper, the control method for a musculoskeletal mechanism simulating a human forearm is discussed. The human muscular system which is basically the multi-feedback system, serves for skillful human motions. The joint's compliance is controllable to regulate the spinal reflex feedback gains of muscle's length and tension. Furthermore, the cooperation of biarticular and uniarticular muscles improve the manipulation ability and the efficiency of energy. Because the tendon-driven system is similar to the human muscular system, we can take these superior characteristics of the human muscular system into our manipulator. Our manipulator has a similar mechanism to the human forearm and has 2 D.O.F motions (flexion, extension and external, internal rotation) by three contending tendon-driven systems. The control method for this manipulator is designed by imitating the human muscular control system based on spinal reflexes. In the controlled system, each wire compliance is adjustable to change the feedback gains of wire length and tension, and the manipulator compliance under the position control can be controlled by the regulation of the wire compliance without measuring the outer force.
Three-axis force sensors based on the principle of pressure pick-up device are characterized by their manufacturability, simple structure, and miniature size. This paper proposes a new six-axis force sensors composed of two three-axis force sensors and connector. This type of sensor is called Twin-Head type Six-Axis Force Sensor or TH force sensor for short. This paper first describes the theoretical basis for the operation of TH force sensors and provides the characteristic matrix connec-ting the load and sensor output vectors. It is shown that six types of three-axis sensor are theoretically conceivable, and proved that while there are twenty-one combinations, ten combinations can result in their characteristic matrices full ranks. An optimal design approch is introduced to obtain the best sensor configuation using Singular Value Decomposition (SVD) technique. We also show a design orientation using practical six-axis force sensor with experimental validation.
This paper presents a new strategy for motion planning of multiple robots as a multi-agent system. The system has a decentralized configuration. All the robots cannot communicate globally at the same time, but some robots can communicate locally and coordinate to solve competition for a public resource. In such system, it is difficult for each robot to plan its motion effectively while considering other robots, because the robots cannot predict motions of other robots as an unknown environment. Therefore, each robot only determines its motion selfishly for itself whicle considering a known environment. In the proposed approach, each robot plans its motion while considering the known environment and using empirical knowledge. The robot considers its unknown environment including other robots in the empirical knowledge. The Genetic Algorithm is applied to optimization of motion planning of each robot. Through iterations, each robot acquires knowledge empirically using fuzzy logic and the system behaves efficient evolutionary. For illustration, this paper deals with path planning of multiple mobile robots and creates simulations.