基于智能PID和扩张状态观测器的姿态控制方法

Attitude Control Method Based on Intelligent PID and Extended State Observer

针对在实际工程中人工设计飞行器姿态控制参数造成工作量大的问题,提出一种基于模型参考自适应控制的神经网络姿态控制方法。首先在模型参考自适应控制的基本框架下,利用BP神经网络的自学习特性,实现PID控制参数的自适应整定。随后设计了两阶线性扩张状态观测器以对控制量进行补偿,并对观测器进行了误差分析。仿真结果表明通过神经网络训练得到的控制参数变化规律和工程实际较为相似,姿态角指令跟踪误差在允许范围内,幅值裕度和相角裕度均符合要求。通过分别在常值干扰和正弦变化干扰下的两阶线性扩张状态观测器和三阶线性扩张状态观测器的仿真对比,发现两阶的观测器更能够降低外界扰动对系统的影响。

A neural network attitude control method based on model reference adaptive control is proposed to solve the problem of heavy workload caused by manual design of aircraft attitude control parameters in practical engineering.Firstly,under the basic framework of model reference adaptive control,the self-learning characteristic of BP neural network is used to realize the adaptive adjusting of PID control parameters.Then,a two-order linear extended state observer is designed to compensate the control variable and the error analysis is implemented for the observer.The simulation results show that the changing law of control parameters obtained by neural network training is similar to the engineering practical parameters,the attitude angle instruction tracking error is within the allowable range,and the amplitude margin and phase angle margin meet the engineering requirements.Through the simulation comparison results between the two-order linear extended state observer and the third-order linear extended state observer under the disturbance of constant value and sinusoidal change,it is verified that the two-order observer can better reduce the influence of external disturbance.