Nov 7 2019
An international team of researchers has contributed research efforts to make unmanned aerial vehicles (UAVs) smarter.
Image Credit: Dmitry Kalinovsky/Shutterstock.com
The team devised a technique to enable several UAVs to follow a formation while they still automatically regulate their own flight requirements.
The outcomes of the study have been reported in IEEE/CAA Journal of Automatica Sinica, a collaborative publication of IEEE and the Chinese Association of Automation.
For a long time, UAVs have been studied for their prospective use in commercial or military applications. This is because they are highly energy-efficient in spite of carrying comparatively large payloads.
The holy grail of such research is to have formations of UAVs that are able to complete missions autonomously with little supervision for the human operator.
Simone Baldi, Study Author and Professor, School of Mathematics, Southeast University
Baldi is also an assistant professor at the Delft Center for Systems and Control.
In this study, Baldi and his colleagues have used a method known as “software-in-the-loop.” This enables the UAV to be directly tested using the autopilot software already loaded onto the UAV. The direct testing will enable the researchers to move faster from the testing stage to real-time flight.
The scientists configure realistic simulation environments for UAVs fitted with adaptive controllers to simulate flight under different environmental conditions. The adaptive controllers can assist the UAVs in understanding their environment, and also adapt during any change in the conditions.
UAVs equipped with adaptive controllers are expected to optimize their flight operation and adapt to environmental changes, such as different loads, faults or wind currents.
Simone Baldi, Study Author and Professor, School of Mathematics, Southeast University
As a next step, the scientists intend to execute real-time flights for groups of UAVs. Baldi stated that very few institutes and companies across the globe have the capability to conduct reliable flights with formations of UAVs.
We built a simulation environment that is able to learn and adapt to the characteristics of each UAV. The aim is to test not only a single UAV but teams of UAVs, each with different features.
Simone Baldi, Study Author and Professor, School of Mathematics, Southeast University
Jun Yang from the Systems Engineering Research Institute at the China State Ship-building Corporation; Ximan Wang and Satish Singh, both from the Delft Center for Systems and Control; and Stefano Fari of the German Aerospace Center, at the Institute of Space Systems are the other contributors to the study. Fari is also affiliated to the Delft Center for Systems and Control and the Politecnico di Milano.
This study was financially supported by the Fundamental Research Funds for the Central Universities, the Special Guiding Funds for Double First-Class, and the Delft Center for Systems and Control.