High-precision gravity data are an important foundation for key national defense technologies such as passive inertial navigation for submarines and underwater obstacle detection. Hybrid sensor systems that combine quantum interferometers and classical accelerometers offer a promising route for high-precision dynamic acceleration measurements. However, their measurement accuracy on moving platforms is severely limited by motion-induced cross-coupling errors, a critical systematic effect where horizontal accelerations spuriously leak into the vertical measurement channel. Traditional methods relying on single-objective optimization of interference fringe quality often lead to overfitting and poor external consistency accuracy. To overcome this limitation, this paper proposes a Multi-Objective Joint Optimization (MOJO) framework for motion-induced dynamic disturbance mitigation. A dual-objective cost function is formulated to simultaneously minimize the atom interferometer's phase noise and the spurious low-frequency correlation between the recovered vertical signal and horizontal platform motion, thereby ensuring the physical consistency of the measurement results with geophysical priors. The framework is validated through a shipborne dynamic gravity survey. Experimental results demonstrate that MOJO improves the external consistency accuracy of the hybrid system from 2.30 mGal to 0.78 mGal, a 66.1% improvement over the uncorrected case and a further 27.8% improvement over conventional single-objective optimization methods. The framework also exhibits superior robustness across various dynamic scenarios.
GONG et al. (Sun,) studied this question.