基于自适应模糊神经网络的反舰导弹非线性控制算法

Nonlinear Control Algorithm Based on Adaptive Fuzzy Neural Network for Anti-ship Missile

提出了一种综合运用动态逆、模糊神经网络和滑模控制的非线性控制方法。首先运用动态逆理论对非线性系统进行近似线性化,利用具有在线学习能力的模糊神经网络来抵消系统的误差,建立了基于自适应模糊神经网络的控制结构,根据李雅普洛夫稳定理论导出了网络权值的自适应调整规则,用滑模控制和鲁棒控制分量保证了系统的鲁棒性。并将该非线性控制算法用于某型侧滑转弯导弹的控制系统设计。仿真结果表明,这种方法能有效消除扰动的影响,提高导弹过载控制系统响应的精度。

A new nonlinear control algorithm using dynamic inversion , fuzzy neural network and sliding mode control was presented. The dynamic inversion theory was used to approximately linearize the nonlinear system, the system error was compensated by capable of on-line learning fuzzy neural network, and based on adaptive fuzzy neural network a control structure was built. A stable weights adaptive adjustment rule was derived by using Lyapunov stability theorem, and the system robustness was guaranteed by using sliding mode control and robustness control component. This control law was applied to onc sides to turn（STT） missile control system, simulation results showed this control law can eliminate the influence from disturbance availably, raise the precision of missile overload control system response.