论文部分内容阅读
采用BP神经网络逼近算法,推导、建立了预定关机点及落点坐标与需要速度间的映射关系,并装订上弹,使关机点附近对应的每一个位置都能够映射出相应的需要速度矢量,利用闭路制导关机及导引方法对导弹实施控制。提出了基于BP神经网络制导的数据制备方法,将大量的弹道数据改为以神经网络形式进行制备,有效地缩小了数据存储空间及弹载计算机的计算时间。仿真结果表明,基于BP神经网路的制导方法能够大大提高导弹的制导精度,相应的诸元数据制备方法能够准确地实现神经网络在弹上的映射功能,为该制导方法在弹上的应用奠定了基础。
The BP neural network approximation algorithm is used to derive and establish the mapping relation between the scheduled shutdown point and the falling point coordinates and the required speed and bind the upper bound such that each position corresponding to the shutdown point can map the corresponding required velocity vector, Guidance and Control of Missile with Closed-Circuit Guidance Shutdown and Guidance Method. A data preparation method based on BP neural network guidance is proposed. A large amount of trajectory data is changed to be prepared in the form of neural network, which effectively reduces the data storage space and the computational time of the onboard computer. The simulation results show that the guidance method based on BP neural network can greatly improve the guidance accuracy of the missile. The corresponding metadata preparation methods can accurately map the neural network on the projectile and lay the foundation for the application of the guidance method in the projectile The foundation.