Distributed Operational Space Formulation of Serial Manipulators

Author(s):  
Kishor D. Bhalerao ◽  
James Critchley ◽  
Denny Oetomo ◽  
Roy Featherstone ◽  
Oussama Khatib

This paper presents a new parallel algorithm for the operational space dynamics of unconstrained serial manipulators, which outperforms contemporary sequential and parallel algorithms in the presence of two or more processors. The method employs a hybrid divide and conquer algorithm (DCA) multibody methodology which brings together the best features of the DCA and fast sequential techniques. The method achieves a logarithmic time complexity (O(log(n)) in the number of degrees of freedom (n) for computing the operational space inertia (Λe) of a serial manipulator in presence of O(n) processors. The paper also addresses the efficient sequential and parallel computation of the dynamically consistent generalized inverse (J¯e) of the task Jacobian, the associated null space projection matrix (Ne), and the joint actuator forces (τnull) which only affect the manipulator posture. The sequential algorithms for computing J¯e, Ne, and τnull are of O(n), O(n2), and O(n) computational complexity, respectively, while the corresponding parallel algorithms are of O(log(n)), O(n), and O(log(n)) time complexity in the presence of O(n) processors.

Robotica ◽  
2021 ◽  
pp. 1-12
Author(s):  
Paolo Di Lillo ◽  
Gianluca Antonelli ◽  
Ciro Natale

SUMMARY Control algorithms of many Degrees-of-Freedom (DOFs) systems based on Inverse Kinematics (IK) or Inverse Dynamics (ID) approaches are two well-known topics of research in robotics. The large number of DOFs allows the design of many concurrent tasks arranged in priorities, that can be solved either at kinematic or dynamic level. This paper investigates the effects of modeling errors in operational space control algorithms with respect to uncertainties affecting knowledge of the dynamic parameters. The effects on the null-space projections and the sources of steady-state errors are investigated. Numerical simulations with on-purpose injected errors are used to validate the thoughts.


Author(s):  
James H. Critchley

A new and efficient form of Featherstone’s multibody Divide and Conquer Algorithm (DCA) is presented. The DCA was the first algorithm to achieve theoretically optimal logarithmic time complexity with a theoretical minimum of parallel computer resources for general problems of multibody dynamics, however the DCA is extremely inefficient in the presence of small to modest parallel computers. The new efficient DCA approach (DCAe) demonstrates that large DCA subsystems can be constructed using fast sequential techniques and realize substantial speed increases in the presence of as few as two parallel processors. Previously the DCA was a tool intended for a future generation of parallel computers, this enhanced version promises practical and competitive performance with the parallel computers of today.


2020 ◽  
Vol 39 (10-11) ◽  
pp. 1239-1258
Author(s):  
Shameek Ganguly ◽  
Oussama Khatib

Multi-surface interactions occur frequently in articulated-rigid-body systems such as robotic manipulators. Real-time prediction of contact-interaction forces is challenging for systems with many degrees of freedom (DOFs) because joint and contact constraints must be enforced simultaneously. While several contact models exist for systems of free rigid bodies, fewer models are available for articulated-body systems. In this paper, we extend the method of Ruspini and Khatib and develop the contact-space resolution (CSR) model by applying the operational space theory of robot manipulation. Through a proper choice of contact-space coordinates, the projected dynamics of the system in the contact space is obtained. We show that the projection into the dynamically consistent null space preserves linear and angular momentum in a subspace of the system dynamics complementary to the joint and contact constraints. Furthermore, we illustrate that a simultaneous collision event between two articulated bodies can be resolved as an equivalent simultaneous collision between two non-articulated rigid bodies through the projected contact-space dynamics. Solving this reduced-dimensional problem is computationally efficient, but determining its accuracy requires physical experimentation. To gain further insights into the theoretical model predictions, we devised an apparatus consisting of colliding 1-, 2-, and 3-DOF articulated bodies where joint motion is recorded with high precision. Results validate that the CSR model accurately predicts the post-collision system state. Moreover, for the first time, we show that the projection of system dynamics into the mutually complementary contact space and null space is a physically verifiable phenomenon in articulated-rigid-body systems.


Author(s):  
James H. Critchley ◽  
Kurt S. Anderson

Optimal time efficient parallel computation methods for large multibody system dynamics are defined and investigated in detail. Comparative observations are made which demonstrate significant deficiencies in operating regions of practical importance and a new parallel algorithm is generated to address them. The new method of Recursive Coordinate Reduction Parallelism (RCRP) outperforms or directly reduces to the fastest general multibody algorithms available for small parallel resources and obtains “O(logk(n))” time complexity in the presence of larger parallel arrays. Performance of this method relative to the Divide and Conquer Algorithm is illustrated with an operations count for the worst case of a multibody chain system.


Author(s):  
Rudranarayan M. Mukherjee

This paper presents a generalization of the divide and conquer algorithm for sensitivity analysis of dynamic multibody systems based on direct differentiation. While similar sensitivity analysis approach has been demonstrated for multi-rigid and multi-flexible systems in tree topologies and a limited set of kinematically closed loop topologies, this paper presents the generalization of these approaches to systems in generalized topologies including many coupled kinematically closed loops. This generalization retains the efficient complexity of the underlying formulations i.e. linear and logarithmic complexity in serial and parallel implementation. Other than the computational efficiency, the advantages of this method include concurrent sensitivity analysis with forward dynamics, no numerical artifacts arising from parametric perturbation and significantly reduced data storage compared to traditional methods. An interesting application of this work in control of multibody systems is discussed.


Author(s):  
James H. Critchley ◽  
Kurt S. Anderson ◽  
Adarsh Binani

A new and efficient form of Featherstone’s multibody divide and conquer algorithm (DCA) is presented and evaluated. The DCA was the first algorithm to achieve theoretically the optimal logarithmic time complexity with a theoretical minimum of parallel computer resources for general problems of multibody dynamics; however, the DCA is extremely inefficient in the presence of small to modest parallel computers. This alternative efficient DCA (DCAe) approach demonstrates that large DCA subsystems can be constructed using fast sequential techniques to realize a substantial increase in speed. The usefulness of the DCAe is directly demonstrated in an application to a four processor workstation and compared with the results from the original DCA and a fast sequential recursive method. Previously the DCA was a tool intended for a future generation of parallel computers; this enhanced version delivers practical and competitive performance with the parallel computers of today.


Author(s):  
James H. Critchley ◽  
Adarsh Binani ◽  
Kurt Anderson

A new and efficient form of Featherstone’s multibody Divide and Conquer Algorithm (DCA) is presented. The DCA was the first algorithm to achieve theoretically optimal logarithmic time complexity with a theoretical minimum of parallel computer resources for general problems of multibody dynamics, however the DCA is extremely inefficient in the presence of small to modest parallel computers. This alternative efficient DCA approach (DCAe) demonstrates that large DCA subsystems can be constructed using fast sequential techniques to realize a substantial increase in speed. The usefullness of the DCAe is directly demonstrated in an application to a four processor workstation and compared with results from the original DCA and a fast sequential recursive method. Previously the DCA was a tool intended for a future generation of parallel computers, this enhanced version delivers practical and competitive performance with the parallel computers of today.


2005 ◽  
Vol 29 (4) ◽  
pp. 679-690 ◽  
Author(s):  
Liguo Huo ◽  
Luc Baron

This paper introduces the concept of functional redundancy of serial manipulators, and presents a new resolution scheme to solve such redundant robotic tasks requiring less than six degrees-of-freedom. Instead of projecting the secondary task onto the null space of the Jacobian matrix in order to take advantage of the redundancy, the twist of end-effector is directly decomposes into two orthogonal subspaces where the main and secondary tasks lie, respectively. The algorithm has shown to be computationally efficient and well suited to solve functionally-redundant robotic tasks, such as arc-welding.


Robotica ◽  
2007 ◽  
Vol 25 (5) ◽  
pp. 511-520 ◽  
Author(s):  
Bojan Nemec ◽  
Leon Žlajpah ◽  
Damir Omrčen

SUMMARYThis paper deals with the stability of null-space velocity control algorithms in extended operational space for redundant robots. We compare the performance of the control algorithm based on the minimal null-space projection and generalized-inverse-based projection into the Jacobian null-space. We show how the null-space projection affects the performance of the null-space tracking algorithm. The results are verified with the simulation and real implementation on a redundant mobile robot composed of 3 degrees of freedom (DOFs) mobile platform and 7-DOF robot arm.


Author(s):  
Amin Kamalzadeh ◽  
Leila Notash

Wire-actuated robot manipulators are generally lighter than other manipulators as actuated wires are used instead of joint actuators. The inverse dynamic modeling of these manipulators is complicated by the existence of multiple kinematic constraints as well as redundancy in actuation. In wire-actuated parallel manipulators with a constraining linkage and in tendon-driven serial manipulators, wires are used to control the joints. In these manipulators, each wire can provide a torque/force on a link about/along its revolute/prismatic passive joint in one direction, as wires only act in tension. Using one wire for each link sometimes does not fully constrain the motion of the link about/along its passive joint. Therefore, a second wire is attached to some links in a “counterbalance” configuration; i.e., the second wire can provide a “complementary” torque/force in the opposite direction of the torque/force produced by the first wire on the link about/along its passive joint. Depending on the end effector trajectory and external force at each instant, one of the mentioned two wires provides the desired direction of torque/force and the other, “counteracting wire,” imposes a “counteracting” torque/force on the link about/along its passive joint. Using more actuators than degrees of freedom (DOF) in the manipulator causes redundancy in actuation, which means that for a unique end effector trajectory and external force, inverse dynamic results (actuator torques/forces) have infinite solutions within a null space of actuator torques/forces. Obtaining a unique result within the null space requires several considerations, such as avoiding negative tensions in wires and decreasing the actuator torques/forces. The purpose of this article is to find a methodology to limit the infinite inverse dynamic solutions to one while the negative wire tensions are avoided and actuator torques/forces are relatively decreased. As explained in this article, by reducing the counteracting wire tensions, other actuator torques/forces are decreased, because a portion of other actuator torques/forces neutralizes the tensions of counteracting wires. A methodology is developed to detect the counteracting wires in real-time and to present the corresponding tensions to a low positive value; i.e., the counteracting wires are “deactivated.” The proposed methodology can be implemented in the inverse dynamic modeling of wire-actuated parallel manipulators with a constraining linkage and tendon-driven serial manipulators via using the Lagrangian method. This methodology can be used to provide optimum actuator torques/forces and avoid negative tensions in actuated wires. The methodology is implemented in the inverse dynamic modeling of a 4-DOF wire-actuated manipulator where there is one degree of actuation redundancy. In the simulation results, the inverse dynamic model based on the proposed methodology is observed to be quite robust in terms of avoiding negative wire tensions by deactivating the right actuated wire.


Sign in / Sign up

Export Citation Format

Share Document