The study of brain state estimation and intervention methods is of great significance for the utility of brain-computer interfaces (BCIs). In this paper, a neuromodulation technology using transcranial direct current stimulation (tDCS) is explored to improve the performance of steady-state visual evoked potential (SSVEP)-based BCIs. The effects of pre-stimulation, sham-tDCS and anodal-tDCS are analyzed through a comparison of the EEG oscillations and fractal component characteristics. In addition, in this study, a novel brain state estimation method is introduced to assess neuromodulation-induced changes in brain arousal for SSVEP-BCIs. The results suggest that tDCS, and anodal-tDCS in particular, can be used to increase SSVEP amplitude and further improve the performance of SSVEP-BCIs. Furthermore, evidence from fractal features further validates that tDCS-based neuromodulation induces an increased level of brain state arousal. The findings of this study provide insights into the improvement of BCI performance based on personal state interventions and provide an objective method for quantitative brain state monitoring that may be used for EEG modeling of SSVEP-BCIs.