High-pressure liquid CO₂ jets possess the characteristics of low-temperature cooling and dry, residue-free impact, which makes this technology particularly suitable for removing hydroxyl-terminated polybutadiene (HTPB) propellant from decommissioned solid rocket motors. However, existing studies lack multi-objective optimization of impact efficiency and CO₂ consumption, which limits their engineering applications and further promotion. In this study, a high-accuracy quadratic Response Surface Methodology (RSM) relating process parameters to damaged volume was established using a Box–Behnken design (BBD) combined with three-dimensional topography scanning. A theoretical model for CO₂ consumption was developed based on the Homogeneous Equilibrium Model (HEM). On this basis, the Non-dominated Sorting Genetic Algorithm II (NSGA-II) was used to obtain the Pareto-optimal set for maximizing propellant damaged volume and minimizing CO₂ consumption. The results indicate that nozzle diameter has the most significant effect on damaged volume and exhibits a strong interaction with jet pressure. The knee-point solution gives a jet pressure of 15.35 MPa, a stand-off distance of 5 mm, and a nozzle diameter of 1.8 mm. Compared with the initial condition, this compromise condition increases the damaged volume by 72% while increasing CO₂ consumption by only 4.9%. Furthermore, the temperature in the impact zone was reduced to a minimum of −92.4 °C, with no thermal accumulation observed. These findings reveal the influence of liquid CO₂ jet process parameters on impact efficiency and CO₂ consumption, providing a theoretical basis and parameter references for its engineering application in the safe removal of propellants from decommissioned solid rocket motors.

Airborne capture net devices mounted on unmanned aerial vehicles (UAVs) provide an effective physical countermeasure against rogue drones. However, conventional deterministic designs neglect parameter uncertainties such as manufacturing tolerances and assembly errors that degrade flexible net deployment dynamics. To resolve this, this study proposes a multi-objective robust design optimization framework integrating a feedforward neural network (FNN) with uncertainty quantification. A lumped-mass dynamic model is formulated; static deployment experiments validate its high fidelity during the initial phase (t ≤ 0.08 s), while subsequent full-deployment and contraction phases are captured via experimentally calibrated numerical extrapolation. To mitigate computational costs, an adaptive FNN surrogate model is developed, reducing single-evaluation times from hours to seconds. Monte Carlo simulations coupled with the non-dominated sorting genetic algorithm II (NSGA-II) are then employed to optimize the mean and standard deviation of deployment and effective distances. Compared with deterministic optimization, the robust solution yields a 23.8% increase in the 95th percentile of deployment distance with only a 1.5% decrease in effective distance. Crucially, the standard deviations of these metrics decrease by 22.5% and 38.5%, respectively, demonstrating significantly attenuated sensitivity to parameter variations and offering a valuable engineering framework for robust airborne interception.