In a DC-microgrid (DC-MG) composed of a power accumulator battery test system (PABTS), owing to the low inertia of DC capacitance, the charging and discharging of a PABTS can easily cause DC-link voltage fluctuations, which may jeopardize the system stability. Hence, a virtual inertia control (VIC) strategy is proposed to suppress these fluctuations and enhance the stability of the DC-MG. The VIC method is realized in a bidirectional gridconnected converter (BGCC), which combines VIC and model predictive control (MPC). The proposed method can provide inertia support during the transient state and enhance the dynamic characteristics of the DC-link voltage. A prediction model is established that uses the variation range of the DC-link voltage as the constraint, and the output of VIC as well as voltage deviations as optimization objectives. The desired DC-link current increment is calculated using the prediction model to change the input DC current reference of the VIC. To validate the effectiveness of the proposed method, hardware-in-the-loop (HIL) experiments are performed, and the results indicate that MPC-VIC is superior to the existing VIC methods in terms of inertia support and the DClink voltage variation suppression of PABTS DC-MGs.