This paper addresses the problem of efficiently computing the motor torques required to drive a manipulator arm in free motion, given the desired trajectory—that is the inverse dynamics problem. It analyzes the high degree of parallelism inherent in the computations and presents two "mathemati cally exact "formulations especially suited to high-speed, highly parallel implementations using VLSI devices. The first method presented is a parallel version of the recent linear Newton-Euler recursive algorithm. The time cost is linear in the number of joints, but the real-time coefficients are re duced by almost two orders of magnitude. The second formu lation reports a new parallel algorithm that shows that it is possible to improve on the linear time dependency. The real time required to perform the calculations increases only as the [log2] of the number of joints. Either formulation is sus ceptible to a systolic pipelined architecture in which complete sets of joint torques emerge at successive intervals of f...

/pdf/parallelism-in-manipulator-dynamics-22yirvqji7.pdf

Parallelism in Manipulator Dynamics

https://lup.lub.lu.se/search/ws/files/6032282/8498224.pdf

Integrated Architecture for Industrial Robot Programming and Control

Industrial robots play a key role in manufacturing systems. Robots are distinguished from other types of machinery mainly on the basis of their programmability and ability to be adaptable to different tasks. The use of computer control to achieve desired flexibility implies that software issues for embedded control systems are central for the applicability and utilization of the equipment. The structure of control systems today, however, limits the applicability of robots, thus leaving many human unfriendly operations to be performed manually.

This thesis takes a problem oriented approach, without enforcing use of formal methods. Considering industrial demands, such as computing efficiency and simple factory-floor operation, a layered system architecture and technical solutions to accomplish it are proposed. A notion of user views is introduced as the basis for definition of the layers; the layers support programming on levels ranging from implementation of motor control and up to end-user programming. An experimental platform, built around industrially available robots, has been developed. Specially developed hardware interfaces and reconfigurations of the original (ABB) system permits control and programming even of the low level motion control.

Run-time efficiency within the proposed open and layered system was achieved by a new concept called actions. Actions are pieces of compiled code that, by use of certain compiling and linking techniques, can be passed as parameters between the layers. The required interplay between application specific programs and built-in motion control could therefore be accomplished. A number of case studies and results from ongoing experimental evaluation indicate that the proposed control system principles are very useful also in an industrial context.

/pdf/industrial-robot-programming-2hjrnvyl6n.pdf

Industrial Robot Programming

A robot manipulator is a force and motion server for a robot. The robot, interpretating sensor information in terms of a world model and a task plan, issues instructions to the manipulator to carry out tasks.
The control of a manipulator first involves motion trajectory generation needed when the manipulator is instructed to move to desired positions. The procedure of generating the trajectory must be flexible and efficient. When the manipulator comes into contact with the environment such as during assembly, it must be able to comply with the geometric constraints presented by the contact in order to perform tasks successfully. The control strategies for motion and compliance are executed in real time by the control computer, which must be powerful enough to carry out the necessary computations.
This thesis first presents an efficient method for manipulator motion planning. Two fundamental modes of motion, Cartesian and joint, are considered and transition between motion segments is uniformly treated to obtain an efficient and simple system. A modified hybrid control method for manipulator compliance is then proposed and implemented. The method overcomes the problems existing in previous approaches such as stiffness control and hybrid control. Finally, a controller architecture is studied to distribute computations into a number of processors to satisfy the computational requirement in a cost-effective manner. The implementation using Intel's single board computers is also discussed. Finally, to demonstrate the system, the motion trajectory and the modified force/motion control scheme are implemented on the controller and a PUMA 260 manipulator controlled from a multi-user VAX/Unix system through an Ethernet interface.

/pdf/design-and-implementation-of-a-robot-force-and-motion-server-587drsg5v7.pdf

Design and implementation of a robot force and motion server

This paper presents an experimental study of implementation and performance evaluation of hardware and software structures on a parallel processing architecture using Inmos transputers. Based on the requirements of real-time control for large-scale aiid complex systems, a control module composed of a transputer, an A/D converter and a D/A converter for measurement data acquisition and actuation is described. In order to demonstrate the effectiveness of the proposed paral le1 processing control architecture, total performance of liardware and software is experimentally discussed with several structural examples. These are sequential processing on a single transputer imd parallel and distributed processing on multiple transputers for one control module, then paral le1 processing control for two control modules is also investigated. Cycle time and numerical computation time of one control module, and synchronization among multiple control modules in different parallel processing environments are evaluated with respect to the amount of data processing. The advantages of paral le1 processing are discussed in comparison with sequential or concurrent processing.

Implementation Of Real-time Control Schemes On A Parallel Processing Architecture Using Transputers

Summary form only given. A new, distributed processing architecture using three microcomputers in a pipeline configuration is proposed for the efficient execution of a broad class of advanced manipulator control algorithms. This design will facilitate the implementation of advanced control strategies in industrial robots at a competitive cost. A simple example illustrates the advantages of the pipeline architecture over existing solutions.

J. Craig

Papers

Microprocessor architecture for advanced robot control