Robot Arm Motion Research Articles

Deformation Control of Elastic Object by Robot Arm - High-Precision Deformation Control by Adaptive Feed-forward Control -

This paper derives the equations of motion of a horizontal three-link robot arm contacting an elastic object, then proposes deformation control for the elastic object by the robot arm. This deformation control involves feedback control of joint angles, feedforward control by robot arm dynamics, and adaptive feedforward control for unknown elastic object dynamics. An adaptive feedforward control law is devised using an online parameter identification algorithm based on the least squares method, zero phase error tracking control theory, and the Jacobian matrix of the robot arm. Numerical simulation and experimental results confirmed that deformation control error by the proposed adaptive control rapidly decreases with the passing of time.

Deforming Control of Elastic Object by Robot Arm. High-Precision Deforming Control by Adaptive Feedforward Control.

In this paper, first, the equations of motion of the horizontal three-link robot arm in contact with an elastic object are derived. Then, the deforming control system of the elastic object by the robot arm is proposed. The deforming control system consists of the feedback control of the joint angles, the feedforward control by the robot arm dynamics and the adaptive feedforward control for the compensation control of the unknown elastic object dynamics. The adaptive feedforward control law is constructed by use of the on line parameter identification algorithm based on the least squares method, the zero phase error tracking control theory and the Jacobian matrix of the robot arm. From the numerical simulation rerults and the experimental results, it is confirmed that the control error in the deforming control by use of the present adaptive feedforward control method becomes small rapidly with the passing of time.

Development of an automated fixture set-up system for inspection

This paper presents an automated fixture set-up for inspection. The system comprises a workplace with magazines, locators and a baseplate. An articulated robot, the SCORBOT-ER VII is used to automate the system. A C-program is developed and used to execute robot arm movements based on user input. The C-program also establishes locating postions for a rectangular workpiece and determines the set-up of the locating fixtures via points. The program is therefore a useful interface to other small-scale expert systems for fixture design and graphical simulators. The set-up of the fixtures can be performed for other complex shapes if locating points are known, and input into the robot controller using the C-program.

衛星塔載ロボットアームと衛星姿勢の協調制御 ロボットアーム動作時の衛星の姿勢安定の保証

When a robot arm is mounted on a satellite to perform tasks, the satellite attitude must be maintained to retain the communication link and to generate electrical power from its solar pannels. It is not realistic to control the total system as one dynamic system, since the number of degrees of freedom becomes large, and the computational requirement for the satellite-mounted computer becomes stringent. The proposed control method used independent control systems for control of the motion of the robot arm and satellite attitude control. The robot arm control system estimates the angular momentum which will be produced by the robot arm motion, and the attitude control system compensates for the disturbance by using the feedforward control. The robot controller also manages the motion plan of the robot arm in order not to disturb the satellite's attitude stability.

Coordinated control of spacecraft attitude and space manipulators

When a robot arm is mounted on a satellite to perform some tasks, the satellite's attitude must be stabilized to retain the communication link and to generate electrical power from solar panels. It is not realistic to control the total system as one dynamic system, since the number of degrees of freedom becomes too large to be handled by state-of-the-art satellite-mounted computers. This paper proposes a coordinated control between the satellite's attitude control system and the robot arm control system. The robot arm control system estimates the angular momentum of the planned robot arm's motion. The satellite's attitude control system will compensate for the reaction by using feed-foward control. The robot arm controller also manages the motion plan of the robot arm in order not to disturb the satellite's attitude stability.

Approximating Functions by Neural Networks: A Constructive Solution in the Uniform Norm

A method for constructively approximating functions in the uniform (i.e., maximal error) norm by successive changes in the weights and number of neurons in a neural network is developed. This is a realization of the approximation results of Cybenko, Hecht-Nielsen, Hornik, Stinchcombe, White, Gallant, Funahashi, Leshno et al., and others. The constructive approximation in the uniform norm is more appropriate for a number of examples, such as robotic arm motion, and stands in contrast with more standard methods, such as back-propagation, which approximate only in the average error norm. Copyright © 1996 Elsevier Science Ltd

Control Experiment of a Flexible Robot Arm Using the Floating Frame Model

The motion of a straight flexible robot arm is described using a floating frame of reference. This approach allows one to determine a dynamic model with the weakest possible inertia coupling. This model is then easily linearized to design a classical optimal controller. A function accounting for the total length of the path and the vibration intensity at the tip is defined. This controller is applied to a laboratory prototype. Experimental results are in good agreement with numerical simulations. Differences between recordings and numerical predictions can be probably attributed to the difficulties in modelling the actuator at low speed.

A Low Cost Accelerometer for Robot Joint Control

The paper describes a theoretical study of the static and dynamic behaviour of a small “one degree of freedom angular accelerometer” to be used in control of robot arm movements under type II closed loop schemes. The operational principle is based on a “fully floated balanced inertial mass” free to rotate around an axis. Exciting by means of an angular acceleration a consequent inertial torque is created deviating the float. A position transducer measures the deviation angle generating a proportional voltage used to drive a torque motor thus restoring the mass. An “analog closed loop-captured mass system” is obtained where the restitution current in adequate scale is the applied angular acceleration.

Stable and fast neurocontroller for robot arm movement

The authors present new learning algorithm schemes using feedback error learning for a neural network model applied to adaptive nonlinear control of a robot arm, namely the QR-WRLS algorithm and its parallel counterpart algorithms. It involves a QR decomposition to transform the system into upper triangular form, and estimation of the neural network weights by a weighted recursive least squares (WRLS) technique. The QR decomposition method, which is known to be numerically stable, is exploited in an algorithm which involves successive applications of a unitary transformation (Givens rotation) directly to the data matrix. The WRLS weight estimation method chosen allows the selection of weighting factors such that each of the linear equations is weighted differently. The QR-WRLS algorithm is shown to provide fast, robust and stable online learning of the dynamic relations necessary for robot control. We show the results of applying these learning schemes with some flexible forgetting strategies to a two-link manipulator. A comparison of their performance with backpropagation algorithm and the recursive prediction error learning algorithm is included.

Development of Robot Control Program with Optimizing Compiler.

Dynamic control is one of the methods which enable quick and accurate motion of robot arms. However, this method has been rarely applied to industrial robots. One reason is that dynamic control has been difficult to be performed real time by using presently available processors since it requires huge amount of computation. Described in this paper is an optimizing compiler which can automatically generate an efficient program for dynamic control computation of robot arms. Source program, which can be developed with an algebraic computation system like REDUCE, is given as symbolic equations. The optimizing compiler transforms the equations into a set of statements which has a possible simple form of computation. The method is based on the factorization algorithms proposed by authors. After that, the optimized statements are encoded by C Language and arranged for an object program. The optimizing compiler is designed according to an object oriented approach. The effectiveness of the optimizing compiler is also shown through experiments.

Two Arms Are Faster Than One

Two basic approaches to incorporate concurrent movements of multiple robot arms for speeding up a given task are iden tified. The first approach is based on defining a high-level motion primitive that encapsulates two properly synchronized operations into one step. The second approach is based on refining (decomposing) and pipelining robot operations. Both approaches are exemplified by a special pick-and-place task formulated as the Tower-of-Hanoi problem. It is shown that, by using no more than three robotic arms, an exponential speed-up in terms of time steps can be achieved. Therefore, when properly arranged, multiple arms can be extremely cost effective in certain applications.

Passive Joint Control of Dynamic Coupling in Mobile Robots

The results of an investigation on the effects of the dynamic interactions between the vehicle and robot arm motions of a mobile robot system are presented. The objective of this work is to determine whether a mobile robot's dynamic coupling effects can be controlled passively. In this article, various spring- damper combinations (passive control systems) are added to a mobile robot system and evaluated in terms of their effec tiveness in controlling the dynamic coupling. The evaluation procedure utilizes a numerical technique (Runge Kutta-Vemer fifth- and sixth-order method) in conjunction with an algorithm based on the differences between the mobile and nonmobile robot systems' models. The algorithm provides a measure of dynamic coupling and stability. The results are presented in a condensed form via application of Simpson's rule on the cor responding deviations between desired and actual trajectories. In conclusion, the use of passive control systems in conjunc tion with relatively simple controllers (available for existing nonmobile robots and vehicle systems) is expected to provide improved system performance.

Grassmann?Cayley algebra and robotics

Plucker coordinates are a suitable way to represent lines in space for purposes of studying instantaneous motions of robot arms. Similarly, we can represent any affine subspace of Euclidean space of any dimension. This can be useful for studying motion spaces of robot arms. In this paper, after we introduce Plucker coordinates and develop a few of their basic properties, we present them in a somewhat more abstract setting, called the Grassmann—Cayley algebra. Several other applications of this algebra are also described.

MASSIVELY PARALLEL IDA* SEARCH

Heuristic search is a fundamental component of Artificial Intelligence applications. Because search routines are frequently also a computational bottleneck, numerous methods have been explored to increase the efficiency of search. Recently, researchers have begun investigating methods of using parallel MIMD and SIMD hardware to speed up the search process. In this paper, we present a massively-parallel SIMD approach to search named MIDA* search. The components of MIDA* include a very fast distribution algorithm which biases the search to one side of the tree, and an incrementally-deepening depthfirst search of all the processors in parallel. We show the results of applying MIDA* to instances of the Fifteen Puzzle problem and to the robot arm motion planning problem. Results reveal an efficiency of 74% and a speedup of 8553 and 492 over serial and 16-processor MIMD algorithms, respectively, when finding a solution to the Fifteen Puzzle problem that is close to optimal.

Position and Force Neuro-Adaptive Control of a Flexible Robot Arm.

This paper describes the position and force neuro-adaptive control of a two-link flexible robot arm. First, the equation of motion of the flexible robot arm consisting of the first rigid link and second flexible link is shown. The end of the second flexible link is constrained by a rgid wall. Second, the position and force neuro-adaptive control system is composed of two-layer neural networks and proportional feedback control loops. Furthermore, from the experimental results and the numerically simulated ones, it is confirmed that the control error of position and force with the neuro-adaptive control is markedly smaller than the control error without the neuro-adaptive control.

Demonstration of a high-fidelity predictive/preview display technique for telerobotic servicing in space

A highly effective predictive/preview display technique for telerobotic servicing in space under several seconds communication time delay has been demonstrated on a large laboratory scale. The technique is based on a high-fidelity calibration procedure that enables a high-fidelity overlay of 3-D graphics robot arm and object models over given 2-D TV camera images of robot arm and objects. To generate robot arm motions, the operator can confidently interact in real time with the graphics models of the robot arm and objects overlaid on an actual camera view of the remote work site. The technique also enables the operator to generate high-fidelity synthetic TV camera views showing motion events that are hidden in a given TV camera view or for which no TV camera views are available. The positioning accuracy achieved by this technique for a zoomed-in camera setting was about +or-5 mm, well within the allowable +or-12 mm error margin at the insertion of a 45cm long tool in the servicing task.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Motion and force control of multiple robotic manipulators

This paper addresses the motion and force control problem of multiple robot arms manipulating a cooperatively held object. A general control paradigm is introduced which decouples the motion and force control problems. For motion control, different control strategies are constructed based on the variables used as the control input in the controller design. There are three natural choices; acceleration of a generalized coordinate, arm tip force vectors, and the joint torques. The first two choices require full model information but produce simple models for the control design problem. The last choice results in a class of relatively model independent control laws by exploiting the Hamiltonian structure of the open loop system. The motion control only determines the joint torque to within a manifold, due to the multiple-arm kinematic constraint. To resolve the nonuniqueness of the joint torques, two methods are introduced. If the arm and object models are available, an optimization can be performed to best allocate the desired and effector control force to the joint actuators. The other possibility is to control the internal force about some set point. It is shown that effective force regulation can be achieved even if little model information is available.

ＰＩＳ制御を用いたフレキシブルロボットアームの位置と力の制御

In this paper, first, the equation of motion of a horizontal two-link flexible robot arm is derived. The flexible robot arm consists of first rigid link, second flexible link and a control system, and the restriction force is applied to the end of the flexible robot arm. Then, the design method of the control system with PIS control (proportional plus integral control of the strain of the flexible link) is showed, and the experimental results are demonstrated. From the experimental results, it is seen that the control performance of the present position and force control system by use of PIS control is excellent. Furthermore, the dynamic numerical simulation results of the system are demonstrated.

Nonlinear Dynamic Modelling of a Revolute-Prismatic Flexible Composite-Material Robot Arm

In this work, a nonlinear dynamic model is derived to study the motion of a planar robot arm consisting of one revolute and one prismatic joint. Both links of the arm are considered to be flexible and are assumed to be constructed from either isotropic conventional metallic materials or anisotropic laminated fibrous composite materials. The model is derived based on the Timoshenko beam theory in order to account for the rotary inertia and shear deformation. In addition, a nonlinear strain-displacement field is implemented to consider the large deformation of the arm. The deflections of the links are discretized by using a shear-deformable beam finite element. The governing equations of motion are derived from Hamilton’s principle. The digital simulation studies examine the combined effects of geometric nonlinearity, rotary inertia, and shear deformation on the arm’s end effector displacements. Furthermore, effects of the fiber’s angle and material orthotropy on the end effector displacements and maximum normal bending stress are studied.

Robotic Grasping of Unknown Objects: A Knowledge-based Approach

In this article we describe a general-purpose robotic grasping system for use in unstructured environments. Using computer vision and a compact set of heuristics, the system automatically generates the robot arm and hand motions reqrsired for grasping an unmodeled object. The utility of such a system is most evident in environ ments where the robot will have to grasp and manipulate a variety of unknown objects, but many of these manipu lation tasks may be relatively simple. Examples of such domains are planetary exploration and astronaut assis tance, undersea salvage and rescue, and nuclear waste site clean-up. This work implements a two-stage model of grasping: stage one is an orientation of the hand and wrist and a ballistic reach toward the object; stage two is hand preshaping and adjustment. Visual features are first extracted from the unmodeled object. These features and their relations are used by an expert system to generate a set of valid reach/grasps for the object. These grasps are then used in driving the robot hand and arm to bring the fingers into contact with the object in the desired configu ration. Experimental results are presented to illustrate the functioning of the system.