A Rapid Decision-Making Method for Heterogeneous Flight Vehicle in Unsaturated Multi-objective Tasks

To address inefficient damage assessment, insufficient use of collateral damage, and the influence of circular error probable (CEP) and target emplacement error probable (TEP) in unsaturated multi-target firepower planning for heterogeneous rockets, a hierarchical firepower planning method was proposed. First, fragment and blast-wave power fields were represented using energy-aggregated grids, and an enhanced conditional diffusion model (PGen-CDM) conditioned on impact angle, impact velocity, burst height, and warhead type was developed to rapidly generate power fields for different warhead types. Second, polynomial chaos (PC) was used to propagate CEP and TEP and calculate the expected damage caused by coordinated multiple aimpoints. Third, targets were grouped according to the maximum damage range; an upper-level weapon-target assignment (WTA) model incorporating collateral damage and dynamic constraint relaxation was established, and lower-level aimpoint optimization (AO) was then performed within each group. Simulation results show that, compared with direct joint optimization, the proposed method improves the total expected damage by 12.79% and reduces the total planning time from 16 224.67 s to 153.83 s, corresponding to a reduction of 99.05%. Under limited ammunition resources, the method can also prioritize the damage requirements of targets with high strike determination.