Multi-stage trajectory planning of dual-pulse missiles considering range safety based on sequential convex programming and artificial neural network

An algorithm based on sequential convex programming and artificial neural networks is proposed to solve the multi-stage trajectory planning problem of dual-pulse missiles considering range safety. Besides nonlinear dynamics and constraints, the dual-pulse missile introduces many discrete optimization variables (such as the ignition time of the second-stage thrust), and the algorithm needs to consider throwing the engine off to a safe location to ensure range safety, which makes trajectory planning for the dual-pulse missile more difficult to solve. In this study, the whole trajectory is first divided into four stages according to the working characteristics of the dual-pulse engine. Second, three new control variables are introduced to realize nonlinear dynamics convexification, and the relaxation technique is used to relax the constraints between the control variables to avoid non-convexity. Then, a range prediction function is designed to predict the landing location of the engine in real time. To improve the real-time prediction speed, an artificial neural network is further used to fit the range prediction function. Finally, an algorithm combining sequential convex optimization and an artificial neural network is proposed to solve the multi-stage trajectory planning problem of dual-pulse missiles accurately and rapidly. By comparing with pseudospectral method, two trajectory planning cases are solved by simulation, and the effectiveness and rapidity of the proposed method are verified.