Game Development Reference
In-Depth Information
X target
X proxy
g s
k s
g d
k d
F haptic
F simul
Figure 16.4 Admittance to impedance adaptation
because in the case of a virtual wall, the position calculated by the simulation can
accurately maintain the contact and avoid interpenetrations. The behaviour of the vir-
tual object is determined by the environment and its position is not controlled directly,
but indirectly by the forces measured by the device. The functioning of the avatar is
thus completely similar to the other objects of the simulation and simply receives an
additional force. Since all the objects have a similar behaviour, the user can change the
avatar without modifying its behaviour type because all that is required to do is choos-
ing a new object on which the external forces are to be applied. This gives a solution
which is quick and flexible for managing the changes in tool, an operation necessary
in certain types of applications. We would also like to mention that no additional
information needs to be added to the environment for interfacing with the device.
16.2.3 Models primitive to object models (PROXY) Principle
The choice of the control pattern - impedance or admittance - of a device is rarely
given to the user. The devices often use an impedance pattern, though we saw in the
previous section that an admittance pattern is more appropriate as it better respects the
environment constraints and avoids interpenetrations. It is thus necessary to simply
ensure that the device is properly controlled in order to eliminate artificial inertia caused
by the delay in coupling or the one resulting from the device's dynamics. However,
there is a method, directly derived from principles that are well known in automation,
which makes it possible to control an interface in impedance in an admittance mode
(Adams & Hannaford, 1998).
The pattern is shown in figure 16.4. From the simulation we want to recover
forces and send the X proxy positions that are compatible with the environment. From
the device, the X target positions are measured and the forces are awaited. The inter-
mediate control loop then has to evaluate the forces returned to the device and the
simulation on the basis of the position measured by the device and the one calculated
by the simulation. The method consists of simply comparing the positions received as
input and deducing the forces that vary according to the difference e between these
two positions (a simple damper spring of stiffness k and coefficient γ can be enough).
Search Nedrilad ::

Custom Search