In designing a control law for mechanical systems there are at least two methods. One is so called 'Model Based Control Approach(MBC)' and the other is so called 'Dynamics Based Control Approach(DBC) '. Basically the author claim that the control system designed based on DBC is robust against modelling error behaves naturally and becomes friendly for human. In case of DBC of manipulator the fact that manipulator has a property of passivity is well known as a very important feature. In case of legged robot passive dynamic walking is important. In this paper at first we show the property of passivity of manipulator is robust in some sense. Next we introduce a design method of controller for a legged robot based on passive dynamic walking sense. Through these topics we would like to say that the DBC is important for controlling mechanical systems.
11:50―12:10
A Linear Coupling Controller for Plate Vibration
B. Liu, Menasco Aerospace Ltd., F. Golnaraghi, G.R. Heppler, University of Waterloo, Canada
A means of designing linear coupling controllers (LCCs) for multi-degree of freedom systems is developed. An LCC is applied to a thin plate partially clamped on one edge and free on all other edges. The LCC developed here is compared to an LCC designed using the energy monitoring algorithm and to a quadratic non-linear controller. The comparison of the frequency responses for an uncontrolled plate and for the three different controlled cases shows that the design algorithm presented here provides control over a wider frequency range.
Marco A. Arteaga, UNAM, Mexico
Most robust control schemes for rigid robots assume velocities measurements to be available. Although it is possible to measure velocities by using tachometers this increases costs and the signals delivered may be contaminated with noise. Since the use of encoders allows to read joint position pretty accurately sometimes it is desirable to estimate joint velocities through an observer. This paper presents a robust scheme designed in conjunction with a linear observer. Uniform ultimate boundedness for the tracking and observation errors are guaranteed.
12:10―12:30
An Approach to Vibration Control by Stereo Vision System in Mobile Manipulator
Goh Hitaka, Toshiyuki Murakami, Kouhei Ohnishi, Keio University, Japan
A mobile manipulator is composed of the vehicle and manipulator part to expand the workspace area. In case of the rough terrain however the tip motion of the mobile manipulator oscillates. To obtain the stable motion by the manipulator it is necessary to suppress the oscillation. To address the above issue this paper describes a novel strategy to estimate the oscillation of the mobile manipulator by the stereo camera mounted on the vehicle. To recognize the oscillation of the mobile manipulator a landmark point fixed in the workspace coordinate is utilized. Then the landmark point is detected by the stereo camera. From the motion of the landmark points on the image plane of the camera the motion of the mobile manipulator can be estimated. Here to describe the vibration of the mobile …
Adaptive Friction Compensation for Industrial Robot Control
Antonio Visioli, Riccardo Adamini, Giovanni Legnani, Università di Brescia, Italy
In this paper we deal with the friction compensation in the model-based trajectory tracking control of an industrial robot manipulator. First it is shown that the variations of the friction term might significantly affect the control performances during the robot operations. Then a simple adaptive scheme is proposed to solve the problem allowing to keep the trajectory tracking errors at a constant low level. Experimental results obtained in a typical industrial environment show the effectiveness of the method and how it is comparable with known neural-network-based techniques.
12:30―12:50
Constraints Identification for Vibration Control of Time-Varying Boundary Conditions Systems
B. Allotta, Scuola Sant’Anna, F. Angioli, M. Rinchi, Univ. Firenze, Italy
The focus of this paper is on modal identification and updating techniques for a mechanical system characterized by time-varying boundary conditions. In these systems it is very difficult to develop effective robust controllers in reason of the large variations of the dynamic model parameters. In most mechanical systems, such as NC machines, model parametric variations are due to changes in the position of a physical constraint. This fact determines continuous changes in system dynamics (system natural frequencies and vibration modes) while we may assume that constraint properties (such as stiffness and damping coefficients) remain unaltered. Nevertheless constraint properties can't be determined analytically so it is essential to perform experimental tests. Assuming that a FEM model of the …
Advanced Model-Based Robot Control in Reis ROBOTstarV
In a research project that was funded by the Ger-man Federal Government Department of Research an advanced robot control method was conceived and realised, in which a dynamic robot model is integrated in the joint control of the robot Reis RV6. By this model-based control method the path deviations are substantially reduced. This leads to a decrease of actuator input values and thus to lower loads at the actuators and to an improved control behaviour of the robot. As a result, both significant improvements in the path accuracy at high path dynamics and substantially reduced loads of motors and gears have been achieved. The new control method gives an important contribution to higher quality and productivity and also to higher service lives in many application cases.
