基于模糊RBF网络的自适应变结构制导律设计

针对空中高速、大机动目标的拦截问题,传统的滑模变结构制导律虽然对外界干扰和参数摄动具有较好的鲁棒性,但是存在视线角速率抖振及参数不易确定等缺点.对此,在变结构控制理论的基础上,提出了三维自适应变结构制导律,将目标加速度视为外界干扰,设计了一种快速趋近律,并利用模糊RBF网络的高效自学习能力对变结构项的增益进行了在线自适应调节.仿真结果表明,该制导律能够有效削弱系统抖振,提高制导精度,并对拦截大机动目标具有很强的自适应性.

Design of adaptive variable structure guidance law based on fuzzy RBFNN

For intercepting high speed and big maneuvering target in the air,although the traditional variable structure guidance law has good robustness for external interference and parameter perturbation,there are also some shortcomings,such as the jitter of line-of-sight angular rate and undeterminable parameters.In view of this,a three-dimensional adaptive variable structure guidance law was presented which regarded the target acceleration as external interference and adjusted the gain of variable structure adaptively online by Fuzzy RBF neural network which has a good self-learning ability.Simulation results indicate that the guidance law can weaken the system jitter effectively,improve the precision of guidance and has a good adaptability for high maneuvering target.