基于神经网络的导弹自适应滑模控制器设计

针对导弹六自由度非线性模型,根据时标分离原理将导弹系统分为快慢不同的四个回路.针对快、慢回路,提出了一种基于神经网络的自适应滑模控制器设计方案.首先分别在快、慢回路中采用反馈线性化实现解耦控制,然后设计基于神经网络的自适应滑模控制器来保证系统鲁棒性及性能,其中神经网络用来逼近系统的不确定性.理论分析及计算机仿真都表明,按照该方法设计的控制器不仅具有较强的鲁棒性,而且保证了闭环系统的渐近稳定性.

Design of Adaptive Sliding Mode Controller for Missiles Based on Neural Networks

Based on the nonlinear six degree-of-freedom missile model,the missile system was separated into four loops according to time-scale separation principle. Taking account of the fast and slow loops,an adaptive sliding mode controller based on the neural networks was proposed. Firstly,the nonlinear system was decoupled according to feedback linearization in both of the fast and slow loops. And then the adaptive sliding mode controller based on neural networks was designed to guarantee robustness and performance,in which the neural network was used to learn the system uncertainties adaptively. Both of the theoretical analysis and computer simulations indicate that not only the robustness of the controller is great,but also the stability of the closed-loop system is guaranteed.