Abstract: Accurate and efficient aerodynamic prediction methods are of great significance in the early stage of aircraft design, especially during the iterative process of scheme development. This paper develops a data-driven vorticity confinement (DVC) method, aiming to achieve accurate and rapid prediction of aerodynamic characteristics with low-resolution grids. This method adds a volumetric force source term to the momentum equation to compensate for numerical dissipation and by integrating fine-grid high-fidelity data with low-fidelity coarse mesh simulations, establishes a functional relationship between the constraint parameters and the local flow field characteristics reflecting the turbulent flow features through the ordinary least squares method, thereby adaptively adjusting the correction intensity. This calibration framework is open to data sources and can be further extended to direct numerical simulation and experimental measurement data, enhancing the applicability of the method in practical engineering problems. This paper realizes the embedded application of the DVC method in the commercial CFD platform STAR-CCM+, and conducts cross-platform consistency verification in combination with the SU2 results to investigate its portability in the engineering computing environment. After training with flat plate boundary layer data, generalization verification was conducted on the cases of incident shock waves and compression corners at high Mach numbers. The results under the low-resolution coarse grid, the DVC method can accurately predict the velocity distribution of the incident shock wave flow field, significantly reducing the velocity profile error at each section; for the compression corner, both the wall pressure coefficient and the velocity distribution are significantly improved: The relative error of the pressure coefficient is reduced from 27.4% to 5.18% (SU2) and 5.36% (STAR-CCM+). The DVC method simplifies the internal structure of the boundary layer, effectively captures the velocity gradient without relying on traditional fine grids, significantly improving the calculation accuracy and efficiency, and provides a novel engineering solution for rapid aerodynamic prediction in the aerospace field.