攻击区是空空导弹作战性能的重要体现,现代空战目标机动量值大,对攻击区计算提出了更高要求。针对此情况,基于传统BP神经网络,设计了改进BP网络,并进一步与插值法相结合,实现目标大机动和不机动条件的攻击区计算。结果表明,导弹攻击大机动和不机动目标时,改进BP网络插值法和改进BP网络均满足攻击区计算精度要求,且前者性能更优。

Launch envelope was an important reflect of air-to-air missile military performance. In modern air battle, target maneuvering value is large, which puts forward higher requirement for launch envelopes calculation. Aiming at this situation, based on traditional back-propagation (BP) neural network, improved BP network was designed. Furthermore, combined with interpolation method, the launch envelopes calculation of target large maneuvering and non maneuvering condition was realized. The results showed that when missile attacked high maneuvering targets or non maneuvering targets, improved BP algorithm interpolation and improved BP network both met accuracy requirements of launch envelopes calculation, and the former had better performance.