Unlike conventional multicopters, aircraft-type drones have wings that allow them to cover long distances effectively. A new algorithm takes advantage of this design, allowing drones to hover using upward traction. Reducing energy consumption by 150 times Compared to normal active flight.

How is this possible?

The most important feature of the algorithm is ability to adapt to ever-changing wind regimes, while maintaining a constant height. This dynamic control system has the excellent ability to independently find optimal points in upward airflows, allowing for a long hover with minimal energy consumption.

De Kroon’s team took an innovative approach to achieve this goal.

• instead of the usual PID controllers (proportional-integral-derivative) they used the method incremental nonlinear dynamic inversion.
• In this way, the engineers were able to provide alignment to the desired values ​​by facilitating the control of the angular acceleration.

It is noteworthy that this control system seamlessly transitions from hover mode to flight mode without the need for manual adjustment, allowing it to adapt seamlessly to changing wind conditions.

The drone independently finds the optimal position in the air

Engineers are integrated to find optimal hover points in the wind field. simulated annealing algorithm. This algorithm systematically examines the wind field to identify areas where updraft balances the rate of descent, ensuring stable flight with minimal engine thrust. Due to the random direction selection, the algorithm effectively finds equilibrium points and significantly reduces energy consumption.

For testing, engineers created a 3D-printed prototype based on the Eclipson Model C radio-controlled aircraft. This drone had a wingspan of 1,100 millimeters and a weight of 716 grams, including the battery. Flight control was carried out with the Pixhawk 4 remote control, supplemented by a speed sensor, a GPS module and an Optitrack optical system for outdoor and indoor flights.

Drone for tests based on Eclipson Model C aircraft / Photo: S. Hwang et al. / arXiv

How does everything work in practice?

The engineers carried out a series of experiments that checked the operation of the algorithm in different conditions.

• In one experiment, they set the airflow rate between 8.5 and 9.8 meters per second. at a fixed angle of inclination.
• Otherwise, the airflow rate remains constant and tilt angle changed.

In both scenarios, the algorithm proved effective by determining a fixed position in the wind field and holding it for more than 25 minutes. Remarkably, the average engine thrust used during these tests was only 0.25 percent of maximum power, indicating a significant increase in energy efficiency.

Testing the new algorithm: watch the video

The team will test the developments in the field

The team’s success in a controlled environment paves the way for more ambitious outdoor trials in the near future. If these results can be replicated in real-world conditions, it could pave the way for a new era in the energy-efficient use of drones in a variety of industries.

Thus, under the leadership of Delft Technical University engineers Guido de Croon developed a revolutionary control algorithmThis allows aircraft-type drones to hover efficiently using natural wind currents. This progress not only improves energy efficiency, but also opens the door to wider missions and applications of drones.