This study explores the effects of driver memory factors and multi-vehicle speed differences on traffic flow, constructing a model that integrates these elements based on the full velocity difference model (FVDM). The stability conditions of the model were derived by its linear stability characteristics. The traffic initiation and braking processes under the model were simulated and compared with the simulation results of FVDM. A numerical simulation employing the tiny perturbation method was conducted to analyze how driver memory and speed variance among vehicles affect the stability of traffic flow. Finally, the model's parameters were calibrated, and its predictive capability for acceleration was assessed using NGSIM data. The results show that driver memory factor slightly undermines traffic flow stability, while multi-vehicle speed differences has positive effects on the stability of traffic flow. Compared to the FVDM, the proposed model reduces the start-up delay and braking delay by 10.0% and 19.0%, respectively, and achieves a higher prediction accuracy with a reduction of the root-mean-square error by 24.3%.