Analysis and optimization of the conducted emissions of an on- board charger for electric vehicles

Battery chargers for electric vehicles are classified as on-board or off-board chargers. Off-board chargers are not constrained by size or weight but introduce additional cost to the infrastructure through the deployment of a high number of charging stations. In order to meet the needs of electric vehicle users in terms of charging availability, on-board chargers that achieve ac/dc conversion are retained. Furthermore, on-board chargers are classified as standalone or integrated systems. By reusing parts of the traction power train for charging, the latter reduces the cost of the charger. Disadvantages of integrated systems include electromagnetic compatibility issues and complex control schemes.This work presents the power quality performance analysis and control optimization of an on-board non-galvanically isolated electric vehicle charger integrated to the traction’s power train. In order to be able to evaluate the high frequency conducted common mode emissions (150 kHz - 30MHz) of a power conversion structure, one needs to develop a good current control scheme that establishes a high-quality low frequency behavior (0 - 2 kHz). Therefore, different aspects related to the power factor correction of the single-phase as well as the three-phase charging configurations are studied: the control scheme for the regulation of the charging power, the displacement power factor correction, the suppression of the grid current harmonics and the active damping of the input filter’s resonance. Two experimental test benches are developed using two different technologies (Silicon IGBTs vs. Silicon Carbide Mosfets). Experimental results are provided.This work also presents a comprehensive approach to modeling the CM and the DM EMI behavior of a power electronics structure. This method is applied to the charger in its single-phase and three-phase configurations. The models allow to evaluate the fluctuating internal nodes and to study the effect of various proposed mitigation solutions on the CM emissions. The models are also developed in the intent of being injected into optimization algorithms for the future design of an optimal EMI filter.