V3I1P30

Learning-Based Midcourse Guidance Law Design for Surface-to-Air Missiles with Velocity Maximization under Angular Constraints

Nguyen Duc Thang^1*

Abstract

This paper presents a learning-based midcourse guidance law for surface-to-air missiles (SAMs) aimed at maximizing missile velocity while satisfying angular constraints during interception missions. Traditional midcourse guidance laws often neglect explicit consideration of angular limitations, such as seeker field-of-view and structural maneuver constraints, leading to suboptimal energy utilization and degraded interception performance. To address this issue, a learning-based guidance framework is proposed, in which a neural-network-assisted control law is designed to approximate the optimal guidance command derived from a constrained optimization problem. The proposed method integrates classical guidance principles with learning-based adaptation to handle nonlinear missile dynamics and angular constraints effectively. Numerical simulations against maneuvering aerial targets demonstrate that the proposed guidance law achieves higher terminal velocity and improved energy efficiency compared with conventional proportional navigation-based midcourse guidance laws.

Keywords:

Surface-to-air missile, midcourse guidance, learning-based control, velocity maximization, angular constraints