Effects of sampling strategy and delay compensation on the dynamic performance of a bidirectional DC–DC converter for electric vehicle applications

Bidirectional DC-DC converters are widely used in electric vehicle (EV) powertrains. The adoption of digital control in such converters introduces inherent delays, particularly computational and PWM delays, which influence system stability and dynamic performance. Various delay mitigation techniques have been proposed; however, their impact on the frequency-domain characteristics and transient performance of bidirectional DC-DC converters has not been analyzed. This paper presents a delay-aware frequency-domain analysis of a digitally controlled bidirectional DC-DC converter, relating sampling strategies to effective control delay and corresponding stability margins. An advanced multi-sampling PWM technique incorporating sample-shift is implemented in this topology, and a comparison is carried out with single sampling and conventional multi-sampling. The frequency-domain results show that, in buck mode, the proposed approach improves the phase margin from 60.89° to 68.64° and the gain margin from 35.52 dB to 42.87 dB. Time-domain results support these observations, the settling time is reduced from 0.059 s to 0.038 s, and the steady-state error during input voltage disturbances is reduced from 7.73 V to 4.26 V in buck mode. Similarly, in boost mode, the response time to the reference step change improves from 0.00789 s to 0.00688 s, and the settling time is reduced from 0.0747 s to 0.0732 s. Experimental validation supports the effectiveness of the proposed approach.