This paper presents a framework to deal with geometric uncertainties in motion planning for the robots in the vicinity of or in contact with obstacles in their environments. Although geometric uncertainties influence the safeness of the trajectories of the robots in close proximity of obstacles, little attention has been paid to the studies of systematic formalization of geometric uncertainties. Proposed framework utilizes a hierarchical representation: an augmented octree. Octree represents 3-dimensional space by the hierarchy of recursively divided cubes. In order to represent the geometric uncertainties in the object boundaries from the probabilistic point of view, a new concept, occupation probability, is introduced. The occupation probability is defined as the probability that the cube is completely occupied or completely vacant; the spatial uncertainties are expressed by occupation probability values in each octree cube element. Robot trajectories are evaluated by their probability of collision with obstacles. A path planning algorithm to search trajectories by calculating the collision probability is also presented for the purpose of showing the applicability of the proposed framework to motion planning problems.
Disturbance-compensation control of a free-flying space robot is considered in this paper. In the method proposed here, disturbances imposed on a nominal dynamic model are estimated and compensated by utilizing a combination of joint torque input and acceleration signals. The dynamics of the equivalent manipulator with a fixed base is treated as the nominal model, noting that the mass of the base satellite is usually much larger than that of the manipulator. The attitude of the base satellite changes because of the reaction force generated by the manipulator. However, the influence of the change is estimated and compensated as a disturbance based on a combination of the joint torque and the acceleration of the end-effector in the operational coordinate frame. Computation of the proposed control method is as simple as the control of a manipulator with a fixed base. The proposed method is robust against unknown disturbances and modeling errors. This method can also be extended to the case of a space robot with multiple manipulators and attitude control of the base satellite. The effectiveness of the method is demonstrated by computer simulations.
A control method of an anthropomorphic redundant master-slave manipulation system is proposed to be applied to advanced human robot systems, e.g., tele-existence. The impedance control method can be clasified into torque-control type and motion-control type. It is proved by both theoreticaly and experimentally that the motion-control type is suitable for the impedance control of manipulators with reduction gear mechanism. Next control methods of redundant unilateral/bilateral manipulators are proposed and their stabilities are discussed. Efficiency of the proposed methods are demonstrated using a hardware robot system.
We propose a flying vehicle hovering control method using fuzzy model rule which is made by human's control knowledge. We solve the stabilizable condition of this fuzzy controller, and we use this stabilizable condition to realize the learning control. This learning control simulates human's training steps. The control knowledge is shown as the pair of “the condition (for example the upper flying or the lower flying)” and “the operation (controller)”. At the learning, each operation is individually trained under its flying condition. We show the actual hovering experimantation results to show the usefulness of this control method and its stabilization study.
It is necessary for an autonomous mobile robot to solve the problem of obstacles avoidance. Potential method is one of superior to solve this problem. And there are some reports which showed the effect of potential method. But there is few discussion about how to make potential field. In this paper, we try to generate potential field by parallel distributed processing with the group of elements “Sensory Points”, and defined the motion of elements. As a result of computer simulation, we confirmed as follows: it is possible to generate the potential field quickly with sensory points, and potential field which we generated is effective to solve the obstacle avoidance problem.
The wheel mechanism is effective for the flat terrain because of its speed and efficiency. The leg mechanism is effective for the irregular terrain because of its flexibility. Until now, several leg-wheel mobile robots which have both benefits have been constructed. But those have several disadvantages. We have constructed a leg-wheel mobile robot called Chariot which has mechanically separated wheels and legs. Chariot has two large wheels at the center of the body and has two legs attached to the body. Chariot can move at a speed of 20km/h on a flat plane. It can rotate around the vertical axis with small radius. It also can get over step and ditch by changing the phases (the leg-support phase and the wheel-support phase) one after the other. In this paper, as the motion utilizing one of advantages of separated leg-wheel mechanism, we describe the cooperational motion of legs and wheels when Chariot goes over step. Pushing the wheel to the step by legs makes the wheel torque be effective and reduces the load of legs. This causes the reduction of the energy consumption.
Characteristics of the parallel link mechanism are discussed in comparison with a conventional serial link mechanism in terms of kinematics, statics, and sigularity problem. In order to design a high performance manipulator the paralle link mechanism is analyzed by using a sigular value decomposition applied to the forward statics. In this analysis a performance index is introduced that indicates how uniformly the mechanism produces force and moment at its end eff ector. A suboptimal manipulator configuration is proposed that has a better performance index than the conventional Stewart platform.
Several papers were published about the application of the sliding mode control to servo motor, in which the robustness of the control law to the plant parameter variations is the main concern, and the chattering, which is inevitably generated by the delay of the control law and cancellation of the disturbance, is not paid so much attention. Authors have proposed a new approach of sliding mode control combined with disturbance observer, which has the advantage in the robustness with the less chattering even at the relatively low sampling frequency. In this paper, using the Lyapunov stability theorem, an optimal design procedure of disturbance observer based sliding mode CP (continuous Pass) control is proposed, and the result is applied to the two axis SCARA robot trajectory control. Through simulations and experiments, it is verified that the proposed approach is the effective design chart of the observer based sliding mode control, which has the advantage in the less chattering with the robustness of sliding mode control. And the effect of the order of disturbance observer was checked by experiments, which resulted in the fact that the higher order of observer is not so helpful.
An autonomous and decentralized robot system, ACTRESS, is being developed as an intelligent robot system, which performs complex tasks individually or cooperatively. In this paper, the concept of ACTRESS is introduced. ACTRESS is consist of various kinds of robotic agents, for example, robots, equipments, and computers. Then, communication framework, between them is proposed. In ACTRESS, communication is indispensable function, therefore communication framework is one of the most important topics in ACTRESS. This communication framework is designed for both achieving reliable communication and exchanging diverse information. In order to analyze and evaluate implementation of communication framework, a communication simulator has been developed, which simulates parallel behavior of robotic agents and communication between them. This simulator is useful for designing and developing the system, especially its communication function. Finally, by using this simulator, the optimal distribution of function is evaluated. Functional distribution is one of the main characteristics of ACTRESS, therefore the configuration of functional distribution among robotic agents should be designed carefully. The simulator is also useful for evaluating optimality of functional distribution from the viewpoint of communication.