Hybrid guidance law for interception of ballistic targets

Abstract

This article develops a new integrated missile guidance scheme against ballistic targets based on optimal control theory and neural network technology, in which an optimal midcourse guidance law is analytically derived to reach a near head-on interception engagement. This enables an aerodynamically controlled missile to successfully intercept a very high-speed target. A multi-layer feedforward neural network is incorporated with proportional navigation guidance in the terminal guidance section to adaptively enhance missile agility and to correct acceleration commands in response to rapid changes in aerodynamics and target manoeuvres. The entire defensible volume in three-dimensional space is characterized and the missile performance robustness is verified.