Dynamic Cascade Spiking Neural Network Supervisory Controller for a Nonplanar Twelve-Rotor UAV

Area of Science:

• Robotics and Control Systems
• Aerospace Engineering
• Artificial Intelligence

Background:

• Environmental variables like wind disturbance impact multi-rotor unmanned aerial vehicle (UAV) trajectory accuracy.
• Conventional multi-rotor designs often exhibit limited yaw control capabilities.

Purpose of the Study:

• To develop an intelligent supervisory neurocontrol system for precise trajectory tracking in a nonplanar twelve-rotor UAV.
• To enhance the UAV's ability to adapt to environmental uncertainties and improve flight performance.

Main Methods:

• Development of a nonplanar twelve-rotor UAV with a characteristic model for controller design.
• Proposal of an intelligent composite controller integrating adaptive sliding-mode feedback control and dynamic cascade spiking neural network (DCSNN) supervisory feedforward control.
• Implementation of weight learning and dynamic cascade structure learning algorithms for network stability and robustness.

Main Results:

• The proposed DCSNN-based intelligent composite controller demonstrated superior trajectory tracking performance.
• Effectiveness was validated through comparative numerical simulations and prototype experiments under outdoor wind disturbance conditions.
• The nonplanar design improved upon conventional multi-rotor yaw movement limitations.

Conclusions:

• The intelligent supervisory neurocontrol system effectively addresses trajectory tracking challenges in twelve-rotor UAVs, particularly under environmental disturbances.
• The combination of adaptive sliding-mode control and DCSNN offers a robust and adaptable solution for UAV flight control.
• The study confirms the practical viability and enhanced performance of the developed nonplanar UAV and its control system.