5 Control Architecture
Owing to the simplicity of the control law and control structure, it has been widely adopted
and in recent times successfully commercialized (Pires 2008b). The main application
is to manually position the robots for non-interaction tasks for instance welding or for
assisting elderly workers with physically demanding tasks (Neto et al. 2009)(Spillner
et al
. 2008).
5.5.2.1 Single manipulator mode
This structure is a local implementation of the compliance control law on each robot. No
signals whatsoever connect the robots together as shown in Figure 5.7. Each manipulator
has its own compliance controller and gravity compensation. The basic function of this
structure, is that it allows the user to move both robots independently from each other only
with the sheer force of his hand. Thus, by activating one or both compliance control blocks,
the user is able to simply position the TCP of each robot manually. Another function
derived from this is imparting to any manipulator a compliant nature during cooperative
tasks. Before such an action is made possible, it is however imperative to change the
controller gain to accommodate the di
fference between the dynamics of a human hand and
that of a rigid object.
Robot 1
commercial
robot controller
Robot 1
Trajectory
Generator
Robot 2
Robot 2 Trajectory
Generator
commercial robot controller
FTS 2
Environment
FTS 1
Compliance
Controller 2
Compliance
Controller 1
Gravity
Compensation 1
Gravity
Compensation 2
Figure 5.7: Control structure facilitating accommodation control in single manipulator
mode
5.5.2.2 Work-piece mode
On the contrary to the latter implementation, this structure is of a centralized nature.
All manipulators are connected to each other through the work-piece signals. The basic
function of this structure is to enable the compliance control law but on the work-piece level.
72
5.5 Control structures
To achieve the latter objective two steps are necessary. The first is to monitor or observe the
forces on the work-piece i.e. the internal loading of the work-piece due to external forces.
A work-piece observer based on the principle discussed in section 5.4.1 is located after
local force monitoring on the manipulators’ TCP. The second step entails the observation
of the work-piece’s posture w.r.t the world frame based on the instantaneous posture of the
manipulators. By coupling this information with the geometrical constraints, the movement
of the work-piece in space will be consistent and will represent the force dissipation
principle discussed previously. As shown in Figure 5.8 the structure implements the same
concept as that of Figure 5.7 but in a centralized way. However, all the components are
work-piece oriented instead of being specific to one or the other robot. It is also important
to note that the work-piece’s trajectory does not exist, since it stems from interaction
control i.e. no trajectory is given or predefined. The work-piece simply reacts to any force
exerted on it and moves away to reduce it.
Robot 1
commercial robot controller
Robot 1 Trajectory
Generator
Robot 2
Robot 2 Trajectory
Generator
FTS 1
WP Trajectory
Generator
WP Compliance
Controller
FTS 2
Environment
WP Force
Observer
WP Gravity
Compensation
Geometrical
Constraints
commercial robot controller
WP
Position
Observer
Do'stlaringiz bilan baham: