论文部分内容阅读
静态映射神经网络和动态递归神经网络是两种重要的神经网络,前者在系统辨识和控制中得到了广泛的研究和应用;后者能够逼近系统的动态过程,具有良好的稳定性和收敛性,文中针对一类非线性系统,采用动态递归神经网络,结合Lyapunov稳定性理论,综合了稳定的自适应控制器,同时给出了神经网络的自学习律.理论分析表明,在模型匹配的情况下,能够保证跟踪误差的收敛性和闭环信号的有界性,并对一类建模误差问题进行了研究.分析与仿真结果均表明,该方法具有较好的自适应性和鲁棒性.
Static mapping neural network and dynamic recurrent neural network are two important neural networks, the former has been widely studied and applied in system identification and control; the latter can approximate the dynamic process of the system, has good stability and convergence, In this paper, a kind of nonlinear system is adopted, and the dynamic adaptive neural network is used in combination with the Lyapunov stability theory to synthesize a stable adaptive controller. At the same time, a self-learning law of neural network is given. Theoretical analysis shows that under the condition of model matching, the convergence of tracking error and the bound of closed-loop signal can be guaranteed, and a class of modeling error problems are studied. The analysis and simulation results show that this method has good adaptability and robustness.