On the Application of Genetic Algorithms in Underwater Explosive Charges Optimization
Titel:
On the Application of Genetic Algorithms in Underwater Explosive Charges Optimization
Auteur:
Motta, A. A. Ebecken, N. F. F.
Verschenen in:
International journal for computational methods in engineering science and mechanics
Paginering:
Jaargang 8 (2007) nr. 1 pagina's 1-9
Jaar:
2007-01
Inhoud:
Mine strings are sets of submarine mines aligned in one direction, usually built to prevent ships and submarines to pass through an area near the coast (implying that the depths considered are limited). They can be activated either by contact (contact mines) or by acoustic, magnetic, or pressure influence (or their combination). In the first kind, a set of horns is assembled in the mine, which will be activated if one or more of these horns are broken (which only happens by contact), and in the other types a proximity fuse generally senses changes in the respective field and, depending on the processing of these variations, the device can be activated. The interesting thing in proximity mines is that, as the signals captured by the sensors are processed, intelligence can be added to them and they can be set to be activated by certain types of targets only—or even by one specific individual target (in this case, all that is necessary is to have a previously obtained pattern available to be compared during the processing). Mine strings are, therefore, sets of submarines mines, each one with an amount of explosive charge, laid in a certain direction and with a pre-determined distance between them. This distance depends on many factors, such as the explosive charge mass (so that one mine does not damage or activate its neighbors), type and characteristics of threat it is desired to prevent to pass through the mined area, etc. It is possible, then, to find a pair “explosive charge mass + distance between mines” that will result in an optimal mine string to oppose a threat or a class of threats. In this work, a method is presented to calculate the optimal pair W-d combining Genetic Algorithms and matrix-free implicit Finite Elements formulation in a parallel environment with distributed memory, thus providing a tool to support mine strings planning (with this information available, it is possible to build an optimal mine string).
Tijdschrift beschrijving