Intelligent Self-Tuning System for Unmanned Aerial Vehicle

Strategic Insights -

Researchers from Universidad de Malaga have developed a novel backstepping intelligent self-tuning system for a multirotor drone. The system uses a backpropagation neural network to optimize energy consumption and minimize the rise time of the drone, while avoiding unnecessary energy consumption due to overshoot. This breakthrough in unmanned aerial vehicle (UAV) control has the potential to improve the performance and efficiency of drones in various applications, from environmental monitoring to search and rescue missions.

Key Takeaways:

  • The proposed system uses a backstepping intelligent self-tuning method to optimize the performance of a multirotor drone, considering the dynamic response of the system, energy consumption, and rise time.
  • The system employs a backpropagation neural network trained with a database obtained using a metaheuristic algorithm to achieve optimal results.
  • Independent tests were conducted to evaluate the system's performance, with the results showing that it fulfilled the expected dynamic response for 95% of the tests and exhibited a dynamic response with minor overshoot and settling time compared to a PID tuned by genetic algorithm.
  • The system was developed by researchers from Universidad de Malaga, with funding support from the Universidad De Malaga and Consejo Nacional De Humanidades, Ciencias Y Tecnologias.
  • The study's findings have significant implications for the development of efficient and effective UAV control systems.

Statistics:

  • 95% of the independent tests evaluated the proposed method as adequately adjusted and fulfilled the expected dynamic response.
  • The system's dynamic response exhibited minor overshoot and settling time compared to a PID tuned by genetic algorithm.
  • The research was supported by the Universidad De Malaga and Consejo Nacional De Humanidades, Ciencias Y Tecnologias.
  • The study's results were published in the Alexandria Engineering Journal, volume 126, pages 70-80, in 2025.

Sources:

  • Universidad de Malaga
  • Consejo Nacional De Humanidades, Ciencias Y Tecnologias
  • Alexandria Engineering Journal (Elsevier) - http://www.journals.elsevier.com/alexandria-engineering-journal/
  • Neural networks and genetic algorithms-based self-adjustment system for a backstepping controller of an unmanned aerial vehicle. Alexandria Engineering Journal, 2025,126():70-80. (DOI: 10.1016/j.aej.2025.04.034)