This paper considers midcourse guidance of air-to-air tactical missiles. In this paper, an on-line suboptimal midcourse guidance law, which is a neural network approximation of the optimal feedback strategy, is proposed to eliminate the need for solving the two-point boundary-value problems in real time. Although the proposed guidance law is an approximation, it is accurate enough to be suitable for midcourse guidance.
Under the assumption that the optimal control signal can be expressed as a function of the states and final conditions, a neural network is trained to extract the information on the mapping in question from the off-line generated optimal trajectories corresponding to various final conditions. The trained neural network is then employed as a real-time feedback guidance scheme.
Computer simulations are used to compare the performance of this guidance law to that of two types of proportional navigation guidance. The neural net guidance law proposed in this paper shows substantial improvements over the other two guidance laws in simulation results.