论文部分内容阅读
本文针对非线性挠性结构的姿态控制,提出了一种基于高阶神经网络及径向基函数网络(RBFN)相结合的网络模型,用于非线性挠性结构的动态系统辨识,以及基于卡尔曼滤波器(EKF)逆算法的控制策略.针对神经网络辨识时的模型误差,提出了一种简单有效的补偿方法,给出了建模误差补偿与未补偿时的仿真结果.仿真得出,该方法具有收敛快,算法简单,并能有效消除建模误差等优点.
Aiming at the attitude control of nonlinear flexible structures, a network model based on high-order neural networks and radial basis function networks (RBFN) is proposed in this paper. It is used in dynamic system identification of nonlinear flexible structures, Control Strategy of Inverse Algorithm for Mann Filter (EKF). Aiming at the model error in neural network identification, a simple and effective compensation method is proposed. The simulation results of modeling error compensation and non-compensation are given. Simulation results show that the method has the advantages of fast convergence, simple algorithm and effective elimination of modeling errors.