*4.4. Some Discussions*

The obtained results verify that the proposed method can provide better steady-state performance and maintain the inherited dynamic capability. Nevertheless, there are several practice-related issues that should be further discussed.

On the one hand, with respect to the equal voltage value, a DC power supply performs the same electrical characteristic as a healthy battery pack installed onboard the EV when the consideration of battery internal resistance variation is ignored, hence, it is quite feasible to conduct the proof-concept prototype with a DC power supply. It should also be noted that this replacement will be ill-suited when the battery pack suffers from internal resistance variation during a long-term deep discharge process. In practice, the higher internal resistance will lead to a relatively lower output voltage, which may yield an unsatisfactory implementation of the V2G system.

On the other hand, when the grid current has a peak value of 8 A, the voltage loss on the filter inductor will be about 18% (with respect to the voltage of the transformer's secondary side) due to the 0.7 Ω internal resistance of the filter inductor. To compensate for the voltage loss, the dc-link voltage must be increased, which in turn results in a large current ripple as observed in the experimental results. For the alleviation of the problems of the high dc-link voltage accompanied with the large current ripple, the filter inductor with lower internal resistance can be theoretically adopted. Moreover, the utilization of LCL filter can be regarded as an alternative solution because it holds the definite merit of a better performance in terms of the electromagnetic interference suppression.
