中文摘要：

针对棱动司的导弹拦截问题，应用CMAC神经网络技术，并结合变结构控制理论．提出一种新型导引律。首先应用变结构控制理论推导出滑模变结构导引律，然后使CAMC对变结构制导律的输入、输出进行学习，以调整CMAC的连接权值，最终使得制导系统的输出完全由CMAC产生。该导引律克服了变结构制导律中变结构强度项不易确定的缺点，具有较强的自学习、自适应能力以及执行简单的优点。仿真结果表明，所提出的导引律和变结构制导律、比例导引相比在脱靶量、拦截时间等方面有了很大的提高。

英文摘要：

A new guidance law is proposed by application of CMAC neural network technology and variable structure control theory for the passive homing guidance. First of all, sliding mode variable structure guidance law is derived from variable structure control theory, then CAMC study the input and output of variable structure guidance law to adjust the weights, finally the output of guidance system is generated completely by CMAC. The guidance law can overcome the disadvantage of variable structure guidance law and has strong self-learning, adaptive capacity, as well as easy to implement. Numerical simulations show that the proposed guidance law yields better performance than the others.