摘要
本文的研究目的在于使制导律既保留最优制导动态性能好、节省能量的优点,同时又对有界机动目标具有良好的鲁棒性。首先,在导弹-目标追逃问题的相对运动学关系的基础上,我们根据制导动态性能好、节省制导能量要求,构造了线性二次型性能指标,利用Hopfield神经网络在线实时求解该最优制导问题,克服了实际中的最优制导难于求解的问题;同时为了在拦截有界机动目标时,保证视线角速率趋于零,又将滑模控制理论引入到制导律的设计中,并利用Lyapunov稳定性理论对该新型导引律的稳定性进行了证明。仿真结果表明该导引律能够对有界机动目标具有较强的鲁棒性,保证了导弹在追逃过程中使视线角速率趋于零,导弹的指令加速度较小。
Based on the relative engagement kinematics relation between missile and target, we constructed a linear quadratic function performance index, according to the design requirement of improving dynamic behavior and saving control energy. Then Hopfield neural network, which is formulated in the form of an efficient parallel computing device, is employed to solve the optimal control problem for missile guidance on-line in real time. At the same time, in order to assure that line of sight （LOS） rate will achieve zero during the engagement process between missile and target, we also introduced the sliding mode control into the design of missile guidance law, and proved the novel guidance law＇ s stability by Lyapunov stability theory. The robustness to target maneuver is achieved, which can reduce the miss distance. Finally, the simulation results were presented to show the validity of the prooosed method.
出处
《宇航学报》
EI
CAS
CSCD
北大核心
2009年第1期220-224,270,共6页
Journal of Astronautics
