带观测器的RBF分数阶PID在四旋翼飞行器中应用

Neural Network Fractional PID Control of Four-rotor Quadrotor Based on Disturbance Observer

针对经典PID算法抗干扰性差、鲁棒性弱的缺点,给出了带干扰观测器的RBF神经网络分数阶PID控制器。干扰观测器一方面可以将外部的扰动以及由模型参数不准确带来的误差引入到输入端并引入补偿,能很好的提高系统的抗干扰能力以及鲁棒性;另一方面干扰观测器模块和RBF神经网络分数阶PID模块两者互不影响。RBF神经网络分数阶PID可以实现分数阶参数的自学习。最后把带有观测器的RBF神经网络分数阶PID应用与四旋翼控制系统中并且进行MATALAB实验。仿真表明:在环境中存在干扰的情况下,带观测器的RBF神经网络分数阶PID控制器能够良好的实现飞行器的姿态控制,同时具有一定的抗干扰能力以及鲁棒性。

Aiming at the shortcomings of the classical PID algorithm, such as poor anti -interference and weak robustness, the RBF neural network fractional PID controller with disturbance observer is given. Disturbance observer can bring the external disturbance and the error caused by the error of the model parameter to the input end and introduce compensation, which can improve the anti - interference ability and robustness of the system. On the other hand, the disturbance observer module and RBF neural network fractional PID module both do not affect each other. RBF neural network fractional order PID can achieve fractional order parameters of self -learning. Finally, the RBF neural network fractional PID with observer is used in the four- rotor control system and the MATALAB experiment is carried out. The simulation results show that the RBF neural network frac- tional PID controller with observer can do well to control the position of aircraft, and has certain anti - interference ability and robustness.