The precision, stability, and bandwidth of accelerometers are vital for inertial navigation and motion control. This study proposes a quantum-classical hybrid accelerometer that integrates a cold atom interferometer (CAI) with a quartz flexible accelerometer (QFA) using an optimized extended Kalman filter algorithm. High-fidelity numerical simulations based on Markov Chain Monte Carlo methods are conducted to model the sensors’ intrinsic noise and environmental vibration under shipborne conditions. A novel vibration-compensation scheme is introduced to stabilize the CAI output through precise control of the Raman laser phase. As shown in Fig 4, simulation results demonstrate that the hybrid system achieves high-precision, high-bandwidth acceleration measurements in dynamic scenarios, effectively eliminating the long-term drift inherent in classical sensors. Performance analysis via Allan deviation (Fig 4(a)) and noise power spectral density (Fig 4(b)) confirms that long-term instability is suppressed. Under normal sailing conditions, the Kalman filter successfully tracks and corrects up to 91.62% of the QFA measurement errors, Fig 4(c d) presented error of the Titan accelerometer and tracked error, while low-pass filtering further mitigates errors induced by vibration noise. Furthermore, the system exhibits excellent robustness during extreme acceleration events, such as ship collisions. This work provides a comprehensive simulation framework for hybrid quantum-classical accelerometers and highlights their potential for high-dynamic inertial navigation applications.