The traditional quadcopter with four actuators is an under-actuated system, cannot achieve decoupling of line motion and angular motion. Quadcopter with tilting rotors can achieve 6-degree-of-freedom decoupling thanks to the mechanical structure where the motor base allows the motor to rotate around the arm axis, which greatly expands the application scenario of the quadcopter. In this paper, the rigid body dynamics model of the quadcopter with tilting rotors is deduced. In order to solve the non-unique of actuator action caused by over-actuated, a new linear control allocate scheme is adopted. A terminal sliding mode attitude controller(TSMC) based on RBF neural network uncertainty compensator is proposed, and the finite time convergence and global asymptotic stability of the closed-loop system are proved. Parameters of TSMC and RBF are optimized by particle swarm intelligence algorithm with two fitness functions. The main advantage of the proposed method is the ability to observe unknown bounded lumped uncertainties accurately and quickly. Simulation experiments demonstrate the effectiveness and robustness of the proposed controller.
展开▼
