PUMA Arm Research Articles

Inverse kinematic solution and singularity avoidance using a deep deterministic policy gradient approach

<p>The robotic arm emerges as a subject of paramount significance within the industrial landscape, particularly in addressing the complexities of its kinematics. A significant research challenge lies in resolving the inverse kinematics of multiple degree of freedom (M-DOF) robotic arms. The inverse kinematics of M-DOF robotic arms pose a challenging problem to resolve, thus it involves consideration of singularities which affect the arm robot movement. This study aims a novel approach utilizing deep reinforcement learning (DRL) to tackle the inverse kinematic problem of the 6-DOF PUMA manipulator as a representative case within the M-DOF manipulator. The research employs Jacobian matrix for the kinematics system that can solve the singularity, and deep deterministic policy gradient (DDPG) as the kinematics solver. This chosen technique offers enhancing speed and ensuring stability. The results of inverse kinematic solution using DDPG were experimentally validated on a 6-DOF PUMA arm robot. The DDPG successfully solves inverse kinematic solution and avoids the singularity with 1,000 episodes and yielding a commendable total reward of 1,018.</p>

Development of 5 DOF Robot Arm -Gripper for sorting and investigating RTM Concepts

Economy of motion in robots assembly lines, microelectronics, medical and space applications has not received much attention. It is however, as challenging and important as efficiency and ergonomics is to human beings in the factory workplace. Production lines with humans are prone to human fatigue. Hence robotic units are replacing human sorters for production line work. Mechanised operations are usually limited to batch processes with identical items taken from a conveyor belt and placed in a bin or stack. A further extension is to use AI programs to enable discrimination of objects and sorting into separate bins or stacks. In the present study a 5 d.o.f. arm kit was assembled by students and used to understand simple robotic concepts like degree of freedom, gripper action, pick and place, palletisation and other topics in a lab syllabus together with online simulators where factory robots are simulated. Basic algorithms are explored by testing with the 5 dof Puma arm. The addition of tactile and visual interfaces for pick and sort are discussed. Sample outputs are illustrated.

Singularity avoidance of a six degree of freedom three dimensional redundant planar manipulator

This paper focuses on the improvement of singularity avoidance of three dimensional planar redundant manipulators by increasing its degrees of freedom without increasing the number of motors controlling the manipulator. Consequently, the method to build a three dimensional planar manipulator with six-degrees of freedom using three motors instead of six is discussed in detail. A comparison of the manipulability index values for the proposed manipulator is made with the manipulability index values of PUMA arm to demonstrate the effectiveness of using the proposed manipulator for singularity avoidance.

Maximum load carrying capacity of mobile manipulators: optimal control approach

SUMMARYIn this paper, finding the maximum load carrying capacity of mobile manipulators for a given two-end-point task is formulated as an optimal control problem. The solution methods of this problem are broadly classified as indirect and direct. This work is based on the indirect solution which solves the optimization problem explicitly. In fixed-base manipulators, the maximum allowable load is limited mainly by their joint actuator capacity constraints. But when the manipulators are mounted on the mobile bases, the redundancy resolution and nonholonomic constraints are added to the problem. The concept of holonomic and nonholonomic constraints is described, and the extended Jacobian matrix and additional kinematic constraints are used to solve the extra DOFs of the system. Using the Pontryagin's minimum principle, optimality conditions for carrying the maximum payload in point-to-point motion are obtained which leads to the bang-bang control. There are some difficulties in satisfying the obtained optimality conditions, so an approach is presented to improve the formulation which leads to the two-point boundary value problem (TPBVP) solvable with available commands in different softwares. Then, an algorithm is developed to find the maximum payload and corresponding optimal path on the basis of the solution of TPBVP. One advantage of the proposed method is obtaining the maximum payload trajectory for every considered objective function. It means that other objectives can be achieved in addition to maximize the payload. For the sake of comparison with previous results in the literature, simulation tests are performed for a two-link wheeled mobile manipulator. The reasonable agreement is observed between the results, and the superiority of the method is illustrated. Then, simulations are performed for a PUMA arm mounted on a linear tracked base and the results are discussed. Finally, the effect of final time on the maximum payload is investigated, and it is shown that the approach presented is also able to solve the time-optimal control problem successfully.

Timed trajectory generation using dynamical systems: Application to a Puma arm

We present an attractor based dynamics that autonomously generates trajectories with stable timing (limit cycle solutions), stably adapted to changing online sensory information. Autonomous differential equations are used to formulate a dynamical layer with either stable fixed points or a stable limit cycle. A neural competitive dynamics switches between these two regimes according to sensorial context and logical conditions. The corresponding movement states are then converted by simple coordinate transformations and an inverse kinematics controller into spatial positions of a robot arm. Movement initiation and termination is entirely sensor driven. In this article, the dynamic architecture was changed in order to cope with unreliable sensor information by including this information in the vector field. We apply this architecture to generate timed trajectories for a Puma arm which must catch a moving ball before it falls over a table, and return to a reference position thereafter. Sensory information is provided by a camera mounted on the ceiling over the robot. A flexible behavior is achieved. Flexibility means that if the sensorial context changes such that the previously generated sequence is no longer adequate, a new sequence of behaviors, depending on the point at which the changed occurred and adequate to the current situation emerges. The evaluation results illustrate the stability and flexibility properties of the dynamical architecture as well as the robustness of the decision-making mechanism implemented.

Evolution of robotic arms.

The foundation of surgical robotics is in the development of the robotic arm. This is a thorough review of the literature on the nature and development of this device with emphasis on surgical applications. We have reviewed the published literature and classified robotic arms by their application: show, industrial application, medical application, etc. There is a definite trend in the manufacture of robotic arms toward more dextrous devices, more degrees-of-freedom, and capabilities beyond the human arm. da Vinci designed the first sophisticated robotic arm in 1495 with four degrees-of-freedom and an analog on-board controller supplying power and programmability. von Kemplen’s chess-playing automaton left arm was quite sophisticated. Unimate introduced the first industrial robotic arm in 1961, it has subsequently evolved into the PUMA arm. In 1963 the Rancho arm was designed; Minsky’s Tentacle arm appeared in 1968, Scheinman’s Stanford arm in 1969, and MIT’s Silver arm in 1974. Aird became the first cyborg human with a robotic arm in 1993. In 2000 Miguel Nicolalis redefined possible man–machine capacity in his work on cerebral implantation in owl-monkeys directly interfacing with robotic arms both locally and at a distance. The robotic arm is the end-effector of robotic systems and currently is the hallmark feature of the da Vinci Surgical System making its entrance into surgical application. But, despite the potential advantages of this computer-controlled master–slave system, robotic arms have definite limitations. Ongoing work in robotics has many potential solutions to the drawbacks of current robotic surgical systems.

Transputer based GA motion control for PUMA robot

A new control system is described for the PUMA 560 industrial robotic manipulator based on transputer networks, where both the hardware and software designs are detailed. A Transputer Interface Board (TIB) establishing a transputer link to the 6503 microprocessors of the PUMA arm joints has been designed, built and tested successfully. In addition to hardware implementation, software testing for this new system had been accomplished. The new system can communicate with the PUMA lower level controller at a much shorter period (i.e. 1.75 ms) than the default 28 ms. Genetic Algorithms are used to plan the PUMA robot minimum-time motion trajectory, which is not possible by the traditional exhaustive search method. Real-time experiments have been carried out based on the new PUMA control platform, and show a very good match with simulations. Controlled by the new system, PUMA perfoms better when it is interfaced with the shorter time period.

Transputer Control System with a GAs Motion Planner for the PUMA560 Industrial Robotic Manipulator

This paper describes a new control system for the PUMA 560 industrial robotic manipulator based on transputer networks, where both the hardware and software designs are detailed. A Transputer Interface Board (ТIB) establishing a transputer link to the 6503 microprocessors of the PUMA arm joints has been designed, built and tested. In addition to hardware implementation, software testing for this new system has been successfully accomplished. A great deal of flexibility can be achieved with this new system, yet without many difficulties, where it can be used as a platform to implement advanced control algorithms and develop sensor-based intelligent robotic structures which need much more computational power. Genetic Algorithms are used to plan the PUMA robot motion trajectory based on the new PUMA control platform, and the motion planner runs concurrently with the controller. Real-time experiments show that PUMA moves much more smoothly along the optimum planned trajectory.

Fast parallel Preconditioned Conjugate Gradient algorithms for robot manipulator dynamics simulation

In this paper fast parallel Preconditioned Conjugate Gradient (PCG) algorithms for robot manipulator forward dynamics, or dynamic simulation, problem are presented. By exploiting the inherent structure of the forward dynamics problem, suitable preconditioners are devised to accelerate the iterations. Also, based on the choice of preconditioners, a modified dynamic formulation is used to speedup both serial and parallel computation of each iteration. The implementation of the parallel algorithms on two interconnected processor arrays is discussed and their computation and communication complexities are analyzed. The simulation results for a Puma Arm are presented to illustrate the effectiveness of the proposed preconditioners. With a faster convergence due to preconditioning and a faster computation of iterations due to parallelization, the developed parallel PCG algorithms represent the fastest alternative for parallel computation of the problem withO(n) processors.

Error-model-based robot calibration using a modified CPC model

Selection of a proper robot kinematic model is a critical step in error-model-based robot calibration. The Denavit-Hartenberg (DH) model exhibits singularities in calibration of robots having consecutive parallel joint axes. The complete and parametrically continuous (CPC) modeling technique is one of the more versatile alternative modeling conventions designated to fit the needs of manipulator calibration. No modeling convention is, however, perfect. One “user-unfriendly” aspect of the CPC model is a condition handling technique needed, when constructing the error model, to avoid model singularities due to the adoption of the direction vectors of the joint axes as link parameters. This paper presents a modification to the CPC model which brings the model closer to the DH model. Rather than using the direction vectors of joint axes, the modified CPC (MCPC) model employs angular parameters to acommodate the required rotations for each link transformation. This modification results in a much simplified error model. The model, like the CPC model, is capable of completely describing the geometry and motion of the manipulator in a reference coordinate frame. Its error model possesses a minimum number of parameters to span the entire geometric error space and it can be made singularity-free by proper selection of the tool axis. This paper presents a calibration study of the PUMA robot using the MCPC model. A moving stereo camera system was employed for end-effector pose measurements. The MCPC error model was then used for kinematic identification. Results on the PUMA arm show that the MCPC performs well for robot calibration. The well-defined structure and user friendliness of the MCPC model may facilitate the implementation of robot calibration techniques on the factory floor.

Implementation of a unified robot kinematics and inverse dynamics algorithm on a DSP chip

The feasibility and performance of implementing kinematics and inverse dynamics algorithms on a DSP chip for real-time robot arm control is investigated. The algorithms include the following modules: forward and inverse kinematics; Jacobian, inverse Jacobian, and Jacobian derivative term; and Newton-Euler inverse dynamics. These modules are unified under a common coordinate system, and then computationally optimized by eliminating the redundancies among the modules. Further optimization is indicated for the PUMA-like arms. The algorithms are implemented on a TI TMS320C30 DSP chip. It is found that the execution time for the entire set of algorithms is about 0.78 ms for a six-degree-of-freedom robot with a spherical wrist, and is about 0.63 ms for a PUMA-specific arm. The communication time between the host PC and the DSP chip is about 0.376 ms. Thus, it is possible to implement a complete Cartesian controller at a 1000 Hz sampling rate. The algorithms have been successfully tested on a PUMA arm with a PC-based advanced controller.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

A New Dynamic Neurocontrol Architecture for Robot Manipulators

Abstract Neural network control has many innovative potentials for fast, accurate, and intelligent adaptive control. A novel neurocontrol architecture is introduced in which an inverse dynamic neural controller in parallel with a linear neurocontroller cooperates to learn through training to control a robot manipulator accurately under different movement speeds and payloads. Compared with the previous type of neurocontrollers using a fixed PD gain controller, the proposed architecture not only eliminates the painful process of gain tuning but also exhibits a superior control performance because the linear controller constantly adapts its gains while moving. Experimental verification of the controller was carried out for the dynamic control of a PUMA arm demonstrating its superior dynamic control performances compared with a conventional PD or another neurocontrol architecture.

The real-time application of VSS controllers: implementation on elements of a 'direct-drive' arm, a Syke robot and a PUMA robot

The results of the application of new controllers on three robot arms is described. The initial application of variable structure control on a two link direct-drive arm is followed by its implementation on two industrial arms, a Syke robot and a PUMA arm. To allow adequate computational power in the robot controller, a separate controller was built for the first arm which was demonstrated to be portable for use on both of the industrial arms.

Robo-Soar: An integration of external interaction, planning, and learning using Soar

This paper reports progress in extending the Soar architecture to tasks that involve interaction with external environments. The tasks are performed using a Puma arm and a camera in a system called Robo-Soar. The tasks require the integration of a variety of capabilities including problem solving with incomplete knowledge, reactivity, planning, guidance from external advice, and learning to improve the efficiency and correctness of problem solving. All of these capabilities are achieved without the addition of special purpose modules or subsystems to Soar.

Variable structure control of a robotic arm in the presence of uncertainty

AbstractBased on the theory of variable structure systems (VSS), control system design for the trajectory control of robotic systems is presented. It is assumed that the parameters of the system are uncertain and unknown frictional torques are acting at the various joints of the arm. For simplicity, the control of a three‐link PUMA arm is considered. For trajectory following, two control laws, CH and Cθ, based on the choice of coordinates of the end effector or joint angles as the controlled outputs, respectively, are derived. It is seen that whereas control Cθ has no singularity, certain singular surfaces arise where feedback elements of CH become infinity. These singular surfaces describe the boundary of the reachable region (workspace). Digital simulation results are presented to show that accurate trajectory following can be accomplished using the control CH or Cθ for large maneuvers of the arm in spite of the uncertainty in the system.

Decentralized control of nonlinear robot manipulators

Robot manipulators have highly nonlinear dynamics. Therefore the control of multi-link robot arms is a challenging and difficult problem. In this paper a nonlinear dynamic model is first presented for an x- axis robot arm. Then a linearized model is obtained for it. The linearized model is then considered as an n- controller large-scale system. The notion of decentralized stabilization has been introduced and applied to the robot arm. Controllers considered are pd, pid and pd plus a feedforward (inverse dynamics). Numerical simulation for decentralized control of a 3-axis puma arm will also be included.

Model Referenced Adaptive Control for an Industrial Manipulator

The increasing application of robotic industrial manipulators for a wide variety of accurate industrial tasks demands an improved manipulator performance over a wide range of operations. The achievement of desirable dynamic performance in manipulator systems consistently for different operating conditions is difficult when using conventional control methods, because op both the inherent geometric nonlinearities or these systems and the dependence o* the system dynamics on the system configuration and characteristics of manipulated objects. In this study, an attempt is made to improve the manipulator performance to obtain desirable characteristics uniformly, irrespective of payload mass and system configuration using Model Referenced Adaptive Control (MRAC) technique. A nonlinear, coupled dynamic model of a three degree off freedom PUMA arm is developed applying Lagrangian Mechanics, and is simplified to facilitate computation of the system states to be used in digital simulation of the adaptive control algorithms. The adaptive control algorithm, MRAC is developed and its stability is investigated. Using the dynamic model developed earlier, the effectiveness or the adaptive control scheme is evaluated via digital simulation of different maneuvers of the robotic arm that represent practical situations encountered in an industrial setting. The system response is compared without and with MRAC implemented. The effectiveness of the MRAC scheme remains uniformly good irrespective of load and system configuration.

Trajectory control of a PUMA robot arm by inversion and servocompensation

Abstract Based on invertibility and functional reproducibility, an approach to the control of the PUMA arm with six rotational degrees of freedom is presented. The control system has two important features: (i) the ability for fast, precise trajectory tracking; and (ii) robustness to uncertainty in the system, such as variable payload, inertia, etc. The control law derived using invertibility gives decoupled, independently controlled responses in each joint angle using torquers. For robustness, a first-order servocompensator is designed for each decoupled inner loop driven by the joint angle error. For smooth joint angle responses, a command generator is constructed that produces the reference trajectories to be tracked. Simulation results are presented to show that rapid, accurate trajectory following is achieved in the closed-loop system in spite of large payload uncertainty.