Concerned with the problem of interceptor midcourse guidance trajectory online planning satisfying multiple constraints, an online midcourse guidance trajectory planning method based on deep reinforcement learning (DRL) is proposed. The Markov decision process (MDP) corresponding to the background of a trajectory planning problem is designed, and the key reward function is composed of the final reward and the negative step feedback reward, which lays the foundation for the interceptor training trajectory planning method in the interactive data of a simulation environment; at the same time, concerned with the problems of unstable learning and training efficiency, a trajectory planning training strategy combined with course learning (CL) and deep deterministic policy gradient (DDPG) is proposed to realize the progressive progression of trajectory planning learning and training from satisfying simple objectives to complex objectives, and improve the convergence of the algorithm. The simulation results show that our method can not only generate the optimal trajectory with good results, but its trajectory generation speed is also more than 10 times faster than the hp pseudo spectral convex method (PSC), and can also resist the error influence mainly caused by random wind interference, which has certain application value and good research prospects.

In this paper, fractional-order calculus theory is used to investigate the geometric law for intercepting an agile target. In order to overcome the challenges presented by the divergence of line of sight rate (LOSR) of proportional navigation (PN), the fractional LOSR is used as a compensated term in the proposed fractional differential geometric guidance law (FDGGL). By adjusting the navigation gain of the FDGGL, the new proposed guidance law can be transformed into the traditional differential geometric guidance law (DGGL) and PN. The average overload and ballistic stability of the FDGGL are analyzed based on the fractional and control theory. Some analytical results about energy consumption and trajectory variation of the FDGGL were obtained. The simulation results shows that, compared with PN and DGGL, the FDGGL has better guidance performance when intercepting different maneuvering targets. INDEX TERMS missile guidance; differential geometric ; fractional; relative motion; command.

To generate the midcourse guidance trajectory for intercepting the high-speed and high maneuvering target, which is a strongly nonlinear and strongly constrained problem, a two-stage convex optimization method is proposed to solve the optimal trajectory quickly. In the first stage, an initial trajectory generation method is proposed, by which the trajectory’s terminal state is close to the terminal position. And the generated trajectory is used as the initial solution of the convex optimization method. In the second stage, the original nonconvex optimization problem is transformed into a convex optimization problem by linearization and relaxation methods and then solved discretely. In the numerical simulation, the effectiveness of the proposed method is verified, and the robustness of the method is verified in different initial and terminal states. Then, several ablation experiments are operated to verify the advantage of the rapid initial trajectory generation method in the first stage. Finally, compared with the gauss pseudospectral method (GPM), the proposed method is proved to be efficient and has the potential of online trajectory generation.