基于RBF神经网络的X型四旋翼飞行器优化控制

X-type Quadrotor Aircraft Control Based on RBF Neural Network PID

为使X型四旋翼无人机系统能够在受到扰动时快速调整状态以适应环境变化,对X型四旋翼无人机进行区别于传统的受力分析,建立动力学数学模型,采用RBF神经网络和PID联合控制方法,依靠神经网络自学习和非线性映射特征实现系统控制参数的动态整定。以MATLAB/Simulink为实验平台,对RBF神经网络PID控制系统和单纯PID控制系统分别进行仿真。实验结果表明,RBF神经网络PID控制比传统的PID控制调整时间更短、控制效果更好,增强了系统自适应性。

In order to realize the purpose that the x-type four-rotor UAV system can adjust the state in time and quickly to adapt to the change of environment when disturbed,different from the traditional four-rotor uav,the force analysis of the x-type four-rotor uav is carried out in this paper,the reasonable dynamic mathematical model is established,and the combined control method of RBF neural network and PID is adopted for its control.The method relies on the self-learning of neural network and the feature of nonlinear map?ping to realize the dynamic tuning of system control parameters.MATLAB/Simulink was used as the experimental platform to simulate the RBF neural network PID control system and the simple PID control system respectively.The results showed that the RBF neural net?work PID control had less adjustment time and better control effect than the traditional PID control,which increased the system's self-adaptability.