Improved active disturbance rejection control for quadrotor attitude stabilization system
An improved active disturbance rejection control(ADRC)technique is proposed to in-crease the control performance and apply to the inner-loop attitude stabilization system.Firstly,in-spired by the activation function properties of deep learning models,a tracking differentiator with simple structure and easy tuning of parameters is developed using the commonly used sigmoid func-tion.Secondly,a fast-sigmoid tracking differentiator is proposed owing to the fact that the conven-tional sigmoid function is computationally intensive.Finally,since the traditional extended state ob-server(ESO)has the problems of high-frequency jitter and inefficient observation convergence,an improved ESO architecture is proposed by improving the fal function and adding the feedback of ve-locity state,and the convergence judgement is performed.Simulation results show that the improved ADRC controller proposed has better tracking and anti-disturbance capability under the complex dis-turbances.
quadrotorsigmoid functionactive disturbance rejection controlfal function
NETL
NSTL
万方数据
