Sea Technology

MAY 2013

The industry's recognized authority for design, engineering and application of equipment and services in the global ocean community

Issue link:

Contents of this Issue


Page 38 of 87

"The goal is for the communications network to be adaptive according to the instantaneous conditions of the acoustic channel and to the capabilities of the moving nodes." clusions or attenuation may also lead to intermittence in the communication links. All these constraints naturally impact on the formation control, but stable and robust formation is still possible. In this case, the coordination of the vehicles is achieved through a centralized algorithm that considers two types of elements: a virtual leader, which coordinates the operations of the vehicles, and followers, which are instructed by the leader on the actions to perform and/or positions to track. The virtual leader generates the position references for each of the followers, and it can be coincident with any of the physical robots (and use the same computational system). The evolution of the formation takes into account the tracking errors of each robot: the larger the individual errors, the slower the evolution of the formation. By construction, the control algorithm ensures that when any individual error is above a given threshold, the formation stops, i.e., the virtual leader and the communicating followers will hold their positions in the formation, waiting for the remaining vehicle(s) to recover its (their) position(s). Complementarily, in the absence of a communication link with one of the robots for a long time, the formation will hold the position, waiting for status information exchange and position recovery of the missing vehicle(s). This is of special importance in the case of acoustic communications, which may suffer from intermittence due to occlusions or other forms of interference. Although centralized, this approach is particularly well-suited for arbitrary trajectories of possibly time-varying formations. Moreover, the stability of the formation is unaffected by possible communication delays. Obviously, the latency of interaction among the robots degrades the overall performance Sea & Sun Marine Tech of the coordinated system, but the algorithm still guarantees stability. The coordination algorithm requires the vehicles to have local robust control laws that ensure each vehicle is able to track its reference with a bounded error. Therefore, MARES, Zarco and Gama integrate a set of methods that makes it possible to send high-level commands or references. This set includes line following, circle following and target tracking. Under static position reference, the targettracking maneuver also allows station keeping. Commands sent to the vehicles thus involve two main felds: the maneuver and the corresponding reference, which may dynamically change over time. This rather simple and limited set of elemental maneuvers provides a suffciently versatile framework for any complex motion of marine vehicles. Furthermore, combinations of the maneuvers can approximate any type of trajectory. For the coordinated motion of the vehicles, the centralized algorithm makes use of the target-tracking maneuver with dynamic reference points to drive the vehicles in a coherent manner. One of the main advantages of this implementation is that the vehicles do not require knowing their trajectory in advance, which opens the possibility for dynamic path generation and timevarying formation geometries. Testing Several tests carried out thus far have provided successful and encouraging results. As a frst trial, conducted in 2008 in a large dam reservoir in the Douro river, about 25 kilometers east of Porto, the MARES AUV was tracked acoustically, and the tracking information was used to guide the Zarco ASV. Even with acoustic tracking errors in the order of 10 meters, the ASV was able to follow the AUV trajectory. Sea & Sun Technology May 2013 / st 39

Articles in this issue

Links on this page

Archives of this issue

view archives of Sea Technology - MAY 2013