This paper deals with a longitudinal vehicle-following system of two vehicles. Follower vehicle knows relative position by processing the image of preceding vehicle. In order to control speed, optimal control method is used and the cost function is concerned with the errors on command space. The object of estimate is differential of optimal control value so as to eliminate steady state error.
We report a developed autonomous mobile robot (AMR) in dynamically changing environment. The AMR's behavior controller is based upon a state-transition scheme, which suits for realizing the AMR behavior control according to sensory information. The network of state-transition, however, becomes very large as behaviors become complicated. We divide the network among multiple tasks in order to suppress increase of complexity in single network. The tasks consists of a supervisor task and functional tasks. A supervisor task watches overall statuses and events, and controls functional tasks. Each functional task controls a specified part of the AMR's behavior. We constructed a real AMR system, on which the behavior control method is applied and it demonstrated in a showroom as a greeter robot.
The low stiffness due to air compressibilty makes it difficult to control a pneumatic robot accurately. In order to improve its control performance, we propose a robust control scheme using the disturbance observer comprising the plant reference model. The disturbance observer estimates both the variations of plant parameters and disturbances based on the control input and the joint angular velocity. The proposed control scheme can decrease the influence of disturbances by means of the feedback of estimated disturbance. Its availability is investigated through simulations and experiments for the robot arms with two joints. The obtained results can be summarized as follows: 1) To construct the disturbance compensation circuit, the static compensator is used to avoid the nonproper problem. As results, the control system can be easily designed. Further, it can attain the nearly equivalent control performance to which from the dynamic compensator. 2) The proposed control scheme can decrease the influence of not only external force and friction but also interaction force between joints. 3) The positioning perfrmance of robot arms can be improved owing to the effect of disturbance observer.
The trajectory tracking control of a two-link robot manipulator with artificial rubber muscle actuators is described by a learning fuzzy controller. The learning fuzzy controller used here is called a multiple fuzzy controller, in which several elemental fuzzy controllers are processed in parallel and the degree of usage of each inferred consequent is determined by using a linear neural network. The main purpose of the paper is to illustrate the effectiveness of the proposed method by some simulations and experiments for a circular path tracking.
In this paper, we propose“a strategy of following another robot”for motion planning of multiple mobile robot system, which gives order to behavior of robots from the viewpoint of angular velocities of robots. In the strategy, one robot (Follower, F) tries to follow another robot ( Leader, L) which is in front of F and goes in the same direction as F. The strategy is implemented to Virtual Impedance Method, which was proposed by authors. Each robot has two kinds of impedance mode: one is avoiding mode, in which each robot moves individually and independently. The other is following mode, in which each follows L by means of keeping constant distance from L. Each robot decide the mode with other robots in real time by using rule-based method. The simulation indicates effectiveness of the proposed strategy from angular velocities of robots.
The present paper argues that kinematic redundancy of manipulators should be positively utilized in terms of impedance control, and proposes a new impedance control method for redundant manipulators. The proposed method can control not only end-effector impedance using one of conventional impedance control methods but also joint impedance which has no effect to end-effector motion of the manipulator. Regulation of the joint impedance enables to specify dynamic behavior of joints for unknown external forces beforehand. By setting impedance of specific joints very large, for example, it becomes possible to suppress the motion of the corresponding joints and reduce joint degrees of freedom of the manipulator substantially. In this paper, firstly, joint impedance controller is incorporated to the end-effector impedance control system, and a sufficient condition about joint impedance for satisfying a given end-effector impedance is derived. Then, the optimal joint impedance corresponding to a given desired joint impedance is analytically derived using the least square method. Finally, computer simulations are performed using a four-joint planar manipulator in order to confirm of the validity of the method.
Motion characteristics of serial mechanisms at a singular point have been analyzed. When a set of joints of a serial mechanism composes an over-constrained mechanism at a singular point, it is possible to drive the end-effector through the singular point along any specified trajectory. An analytic method to determine joint displacements and velocities at the singular point has been developed. It is shown that the joint displacements and velocities can be determined by the end-effector velocity and acceleration, respectively, in order to drive the end-effector throgh the singular point.
This paper describes the measuring method of the soil and soil model for automated excavation. The measuring method of the soil uses reactive force data which is detected when the manipulator touches the surface of the soil and position data of end effector. From these data, the shape and the hardness of the soil are estimated. The soil model predicts the reactive force of cutting. The cutting process is devided into two stages ; the initial state and the steady state. Different models are adapted to these stages. Experiments are made to show the effectiveness of the model.
This paper describes a method to make a wheeled mobile robot follow a given path just like a driver would decelerate the speed of a car before a corner and accelerate it on the straight course of the road. Two kinds of velocity vectors are introduced to determine an optimal direction and speed of the robot in order to follow the path. The velocity defined as a function of curvature of the path is mainly utilized to change the speed of the robot along the path. The other velocity given as the tangent of a virtual potential function produces an attraction force toward the path. Path tracking control of the robot is formulated with current and future sets of these velocity vectors in a frame of the digital preview and predictive control. By this method, the robust tracking performance is achieved.
In this paper, we consider about a non-linear control scheme of a legged robot called Emu which we are going to develop. The main purpose of this paper is to ensure a kind of stability during the Emu standing-the Emu standing -up or sitting-down very slowly. The important point of our consideration is the following ; (a) we treat the system as a slowly varying system and utilize a concept of frozen system, (b) we applied the exact linearization mehod to the frozen system. The stability is guaranteed by Lyapunov's Theory. The effectiveness of our method is shown by a simulation.