Information fusion-based three-dimensional two-stage optimal cooperative predictive guidance law

Abstract

A three-dimensional (3D) two-stage optimal cooperative predictive guidance law based on information fusion theory is proposed to intercept a highly maneuverable target. This guidance law integrates an Adaptive Predictive Horizon Nonlinear Model Predictive Control (ANMPC) method and a Distributed Anti-Saturation Extended State Observer (DASESO) algorithm. The combined approach aims to minimize the three-dimensional nonlinear zero-effort-miss. By considering zero-effort-miss, the guidance process can be divided into two stages: rapid convergence to maintain state and fine-tuning. The two-stage method allows missile fuel and computational resources to be allocated more targetedly. In the guidance process, a two-stage switching distance is established based on the initial conditions of the missiles and the target. The DASESO algorithm cooperatively processes information from each missile to accurately estimate the target's maneuvering acceleration. Subsequently, an ANMPC method is employed to generate the optimal guidance command. Additionally, the convexity of the optimization problem has been analyzed to ensure its feasibility in practical applications. Through numerical simulation and comparative analysis, it has been proven that this method can effectively intercept and optimize fuel use, even when dealing with highly maneuverable targets with maneuverability similar to that of the missiles.