A smooth robust control approach to cooperation of multiple robot manipulators

Abstract

Byfilly utilizing the properties of the nonlinear dynamics of the system of the multiple cooperating robots with a commonly held object, we propose in this paper a smooth robust controller for trajectory tracking control of the object as well as control of the squeeze force exerted on the object The controller is robust with respect to unknown robot link masses, unknown object mass and unknown joint stick-slip friction. Moreover, fhe need to measure the joint,acceleration or calculate the derivative of the contact force is eliminated in the proposed control design. Unlike other robust controllers, rhe proposed controller is continuously diferentiable, thus ensures smooth control input. Lyapunov's direct method has beenemployed to prove the global asymptotic stability of the closed-loop system. A numerical example of the proposed controller applied to a tworobot system is presented as an illustration.

Observer‐based adaptive control of cooperative multiple manipulators using the Mastroianni operators as uncertainty approximator

“…To achieve the system dynamics, first, the following assumptions are considered. 28,29 Assumption 1. Non-redundancy is considered for each arm.…”

Observer based versus Non-Observer based Adaptive Control of Electrically Driven Cooperative Manipulators using q-analogue of Bernstein-Schurer-Stancu operator as Uncertainty approximator

“…In this figure, ei ei ei ei O X Y Z is the coordinate frame of the ith manipulator which is placed at its grasp point. Now, consider the following assumptions [40][41].…”

Impact model and control of two multi-DOF cooperating manipulators

“…Throughout the remainder of the paper, assume that the functions F (1; 1); h h h 2 (1; 1) and the reference values F r ; 8 8 8 r have been chosen so that q q q a;r and q q q b;r can be uniquely computed as functions of Fr and 8 8 8 r ; at least locally, from (6) as q q q a;r = p p p qa (Fr; 8 8 8r); q q q b;r = p p p qb (Fr; 8 8 8r): (15) Example: In the previously considered case study, it is easy to see that q 1;r and q 2;r can be locally expressed as functions of F r and yr as q 2;r = arccos C + F r 0 Kr The following preliminary feedbacks, each one only using the measurements of the generalized variables of the relative robot, are introduced for the compensation of the gravity forces:…”

“…All the schemes already seen are based on the definition of a controller for each manipulator involved; on the contrary, other schemes (e.g., [15], [22]) consider the set of cooperating robots as a whole system, and a unique controller is defined; in particular, in the solution proposed in [22], a general control paradigm is introduced that decouples the motion and force control problems with reference to the entire robotic system, constituted by multiple robot arms manipulating a commonly held object.…”

“…Various controllers for cooperating multiple robots have been proposed during recent years (see, e.g., [1], [5], [8], [11], [15], [16], and [22]). In several applications, the imposed tasks require a simultaneous control of force and position, as the interaction with another robot or an object to be manipulated is involved, whereas only in few cases the assignment of a particular reference contact force is not needed; for example, in [5] the transport of an object is performed by two cooperating manipulators, by means of a controller constituted for each manipulator by an off-line calculated feedforward part (which includes the compensation of the robot and the object dynamics), and by an on-line computed PID feedback part, plus a possible additional feedforward term for friction compensation, but no force control is present.…”