Self-organised UAV swarms
Tzi-Chieh Chi, John A. Page, Haoyang Cheng, John Ingve Olsen · 2013
The interaction of different search vehicles can be closely associated with the behaviour of natural phenomena. Such examples include foraging ants, pollen-searching honey bees and starlings pre-roosting. The ability to better understand the behaviour and reaction of identical agents led us to investigate the effects of varying each agent's characteristics and provide a way of improving search control and coordination. Evolutionary computation has been investigated since the 1950s and has since been applied to many optimisation problems including blended wing design, power grid optimisation and the study of biological models (growth and development of populations). Detailed analysis of genetic algorithms will be presented in this paper as a powerful multi-parameter function optimisation with simple computational procedures saving time and resources. One of the best ways to solve optimisation problems is to visually simulate the environment as it offers a unique insight into behaviours that can be difficult to investigate through classical mathematical analysis. An example of this insight can be found in the behaviour of swarms consisting of manned and/or unmanned aerial vehicles and ships deployed to undertake complex SAT (Chi, 2012).