Modeling and control approach to a distinctive quadrotor helicopter

Area of Science:

• Robotics
• Control Systems Engineering
• Artificial Intelligence

Background:

• Quadrotor helicopters present unique modeling and control challenges due to complex nonlinear dynamics, inaccurate parameters, and asymmetrical structures.
• Traditional control methods often struggle with the inherent complexities of non-ideal quadrotor configurations.

Purpose of the Study:

• To develop and validate a novel modeling and control strategy for complex quadrotor helicopters.
• To address the nonlinear dynamics and stabilization issues in uniquely configured quadrotors.

Main Methods:

• A Multi-Input Multi-Output (MIMO) Radial Basis Function (RBF) neural network-based state-dependent ARX (RBF-ARX) model was employed to represent the quadrotor's nonlinear dynamics.
• A MIMO RBF-ARX model-based global Linear Quadratic Regulator (LQR) controller was designed for attitude stabilization.

Main Results:

• The proposed MIMO RBF-ARX model-based control approach demonstrated superior performance in stabilizing the quadrotor's attitude.
• Comparative analysis confirmed the effectiveness of the novel method over physical model-based LQR and gain scheduling LQR controllers.

Conclusions:

• The MIMO RBF-ARX modeling and control methodology is validated as an effective approach for quadrotors with complex nonlinearities.
• This study confirms the robustness and superiority of the proposed control strategy for challenging quadrotor systems.