A method of distributed strategymaking in multiple autonomous system is presented in this paper. Every robot in the system has several tactics to be selected, and set of probabilities to select tactics are called “strategy”. The algorithm to make strategy in this paper is iteration of four parts as follows: (1) select and try a certain tactic by means of the strategy, (2) get payoff value for applied tactic from environment, (3) revise the estimated payoff value of tactics, (4) revise the strategy by using (3) . Effectiveness of the proposed method is verified by path selecting simulation of multiple mobile robots. Convergence to optimal solution in the proposed method is discussed by using pay-off matrices in game theory.
A grinding robot system for casted pieces has been developed. The tasted piece grinding requires it to grind the curved surface in narrow area without burning and to follow the casted surface having no precise shape. According to these requisities, the system has been developed with devices which control grinding pressure through a model based compliance control and generate grinding path automatically after getting some design parameters of the work. Next, this paper describes a gyro-moment compensated compliance control. The pot shaped grinder used in this system has large angular momentum, so a gyromoment occurs on it due to robot motion. As the force sensor on the robot detects this moment as an external moment, the robot cannot generate suitable mechanical compliance for the tool. Therefore, we have introduced a new method which can compensate model based compliance control for any effects caused by gyro moment. Finally, we have proved the effectiveness of the system compared with human grinding work. This system is now applied to a manufacturing process of impellers and turbine blades.
Analysing character of object's motion under constraints is important to understand the mechanism of assembly tasks. Such analyses have been studied to solve the contact state transition problem. In those studies, the state of a manipulated object is often represented only by a combination of contacts between the manipulated object and objects in the environment. However, such a representation does not describe all the feature of motion in environments with concave or multiple objects. This paper points out some defects in traditional representation and suggest a solution addressing to the problem.
In this paper we discuss the insertion task of a flexible beam into a hole when the friction force is large enough to buckle the beam. We analyze the buckling shape of the beam and external forces and moments loaded on the beam during the insertion task, under the condition that the potential energy of the beam is minimized. Then, we find that the buckling shape has two buckling modes and the modes are changed by varying the orientation angle of the end of the beam. We propose a strategy of performing the insertion task, based on an analysis of the relation between the external forces and friction coefficient. Finally we present experimental results to confirm our analysis.
A digital nonlinear control of a 3-link direct drive robot manipulator by means of a real time preview virtual desired signal is proposed. The digital nonlinear control is based on the principle that only the acceleration changes at the moment when the input torque changes. A linear optimal control system is designed on the basis of this principle. The real time preview virtual desired signal is the virtual one for improving tracking characteristics for the original desired signal. It is designed considering control characteristics of the already designed digital nonlinear control system and it is derived at real time by feedback of state variables. The real time preview virtual desired signal that means the real time virtual desired signal with preview feedforward compensation is also designed and its characteristics are also studied. The digital nonlinear control system with this real time virtual desired signal is applied to the 3-link direct drive robot manipulator. And frequency characteristics, system responses for circle and square desired orbits are examind in experiment and these effectiveness is confirmed.
In order to evaluate basic performance of a manipulator quantitatively, the manipulability and the dynamic-manipulability are generally used. These concepts were defined for a ground fixed manipulator firstly, and then, the manipulability was applied to a free flying space robot manipulator by using the generalized Jacobian matrix. However, these study for an unconstrained underwater robot have not be carried out, because all momentum of this system is not consevative. In this paper, the dynamic manipulability of an unconstrained underwater robot is defined in the frequency range, by linearizing the kinematic equation and the equation of motion and extending the concept of the accelerance of motion. In this definition of the dynamic manipulability, it was shown that the mapping matrix from driving torque to acceleration of a hand, which is equal to accelerance of a robot system, includes ground-fixed and space robot manipulator dynamics. From the results obtained by the numerical calculations, it was made clear that the dynamic manipulability of an unconstrained underwater robot depends on the frequency of the driving torque, and it is not ellipsoid perfectly, and its measures are smaller than that of space robot because of fluid force acting on the robot.
This paper is concerned with an efficient formulation of a planar variable geometry truss (VGT) . Since the VGT has many topological closed-loops, the number of generalized coordinates is less than that of the members. By using a Lagrangian method, an efficient formulation of the inverse dynamics is derived, because the derived equations of motion have a recursive form. Furthermore, the equations have contained the geometric constraints due to the topological closed-loops. The proposed method needs a less number of calculations in comparison with that of the Newton-Euler method. Effectiveness of the proposed method for inverse dynamics is examined by a numerical simulation.
The TOshiba Modular Manipulator System, TOMMS, consists of joint modules, link modules, and a control unit with a joystick. As to the trial manufacturing, a manipulator with 3 d. o. f. is assembled using three joint modules and optional link modules into any desired configuration and shape, for instance, a horizontal type and a vertical type. The assembled manipulator is connected to the control unit, and the position of the end tip of the manipulator is controlled using the joystick without special handling. There is only one type of joint module and link module. There are three input ports and two output ports on the joint module. The distance between the fore side and the back side of the link module is adjustable. The Jacobian matrix is applied to the control software. Control experiments were carried out and the efficiency of the design concept of TOMMS for mechanical hardware and control software was confirmed.
A control method for a hyper-redundant multijoint manipulator installed on a powerful slider is proposed in this paper. We apply an extension of our Moray Drive control approach, which we introduced earlier. This extension we call as 2 DOF Moray Drive. The 2 DOF Moray Drive derogates the control problem of the hyper-redundant degrees-of-freedom (DOFs) to the simple one of 2 DOFs. Therein, one of the variables signifies the linear combination of the objective initial and final postures of the arm, while the another one expresses the amount of the pull-out-rate, when pulling the arm out of the housing slider. Further on, the optimized motion of the 2 DOF Moray Drive is generated by using the Fourier series expansion method and the culculation convergence method of steepest ascent. The characteristics of the proposed control technique are also demonstrated by computer simulations, including the motion with different gravity conditions, as well as the motion with fluid resistance from water.
In the end point trajectory control of flexible manipulators, a problem of stability arises. This stability is related to the internal stability, since the elastic variables become unstable in the end point control. Some problems have been existed in the end point trajectory control of flexible manipulators considering the internal stabilization. In this paper, first we model a 2 d.o.f. flexible manipulator based on a finit-dimensional order model. Second, we construct a dynamic trajectory controller for the flexible manipulator considering the internal stabilization, and state the problems in controllers of flexible manipulators. Third, we propose a controller using a macro-micro manipulator system which has a small rigid arm at the tip of the flexible manipulator, and show that the problems can be solved by using the macro-micro manipulator system. Lastly, Simulation and experimental results are shown.
This paper proposes a 2-D shape reconstruction method from multi-viewpoint images. It is based on a principle that the shape of an object model is changed so that the projections from the model match with the actual images. This image matching scheme realizes an object shape reconstruction with fusing two kinds of information, i.e., object ridge positions and surface orientations that reflect on edges and shading in images, respectively. Stereo, silhouette contours and shape from shading have the characteristics complement to each other, and the proposed method succeeds them in advantages such as robustness and accuracy. We consider a polygonal prism as a 2-D target object to simplify the problem. Polygon model are treated as a 2-D object model and deformed by a dynamic programming optimization under the condition that the light sources and the camera positions are known.
We believe that personal robots will be used in our office and home like current personal computers in the near future. In this paper, we discuss a user-friendly interface design for personal robots. Then we consider features of personal robots and their environments, and propose a new user interface concept called Active Interface for humanrobot interaction. Active Interface keeps a system in an advantageous condition by using not only users' explicit inputs but also users' implicit inputs and the external information from various sensors. To show the effectiveness of Active Interface concept, we design and implement a speech dialogue system: Chaser based on Active Interface. From the experiments of Chaser, we ascertained the efficiency of Active Interface.