06.02.2013 Views

Underwater Robots - Gianluca Antonelli.pdf

Underwater Robots - Gianluca Antonelli.pdf

Underwater Robots - Gianluca Antonelli.pdf

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

9.2 Kinematic Control of AUVs 231<br />

the lower-priority (secondary) task is further projected onto the null space of<br />

the higher-priority (primary) task soastoremove those velocity components<br />

that would conflict with it.<br />

Adifferent technique to manage conflicting tasks inthe framework of the<br />

task priority approach has been proposed in [29] (see Section 6.6), where a<br />

fuzzy inference system is used tomerge multiple secondary tasks given to<br />

underwater vehicle-manipulator systems.<br />

Obstacle Avoidance Strategy<br />

In [50] obstacle avoidance has been developed in the frame of task priority<br />

inverse kinematics following the approach of Y. Nakamura [210].<br />

In the following, based on the work developed in 1991 at the Università<br />

degli Studi Napoli Federico II for industrial manipulators [76], the obstacle<br />

avoidance is chosen asthe primary task by selecting:<br />

⎡ ⎤<br />

σ o =<br />

⎢<br />

⎣<br />

� η 1 − η o �<br />

.<br />

� η n − η o �<br />

⎥<br />

⎦ (9.9)<br />

where η o is the coordinate vector of an obstacle, and the other tasks as<br />

secondary tasks. Defining as r i the versor of η i − η o it can be recognized that<br />

the pseudoinverse of J o ∈ IR n × 3 n is simply given by:<br />

J † o =<br />

⎡<br />

⎢<br />

⎣<br />

r 1<br />

0 3 × 1<br />

. . .<br />

0 3 × 1<br />

r n<br />

⎤<br />

⎥<br />

⎦ (9.10)<br />

and the null-space projection matrix, considering the planar case for simplicity,<br />

is given by:<br />

⎡<br />

⎤<br />

⎢<br />

N o = ⎢<br />

⎣<br />

1 − r 2 1 ,x<br />

− r 1 ,xr 1 ,y<br />

O 2 × 2<br />

− r 1 ,xr 1 ,y<br />

1 − r 2 1 ,y<br />

. . .<br />

1 − r 2 n,x<br />

− r n,xr n,y<br />

O 2 × 2<br />

− r n,xr n,y<br />

1 − r 2 n,y<br />

In this approach, the obstacle avoidance task is also used for each vehicle<br />

with respect to the other vehicles ofthe platoon. This prevents from the<br />

possible occurrence of accidents during, e.g., formation changing or obstacle<br />

avoidance. In case of simultaneous proximity ofanobstacle and another vehicle,<br />

the safer behavior is obtained by projecting the secondary-task velocities<br />

in the null space of both the obstacle avoidance tasks. This will simply<br />

zero the required velocity for the vehicle that is commanded to go closer to<br />

another vehicle and the obstacle atthe same time.<br />

The developed theory consider holonomic vehicle, in[49] nonholonomic<br />

vehiclesare explicitlyconsideredascomposing the platoon.Recently,in[278],<br />

⎥<br />

⎦ .

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!