以经典控制的弹体闭环回路作为受控对象，研究了基于神经元的导弹自适应控制器。以整体式冲压发动机超音速空地导弹为算例，进行了末制导段弹道仿真。仿真结果表明所设计的控制规律实时性强、抑制干扰性能好，且结构简单，易于工程实现。 Taking the classical closed-loop loop control of the projectile as a controlled object, the neuron-based missile adaptive controller is studied. Taking an integrated ramjet supersonic SAM as an example, the ballistic simulation of the final guidance stage was carried out. The simulation results show that the designed control law has strong real-time performance, good disturbance rejection performance, simple structure and easy implementation.