DC-DC conversion in electric vehicle chargers can require a single or multiple stages depending on both electrical architecture and amount of processed power. When isolation is provided by an output stage, a unidirectional isolated DC-DC converter is required. Among the possible candidates, the LLC resonant converter and the phase shifted full-bridge (FSFB) converter show several advantages which has made them preferred choices in recent literature. This paper compares the efficiency of both converters on equal basis under the variable load conditions imposed by a cycle of the conventional constant-current constant-voltage (CC-CV) charging protocol. To support the study, scale-down experimental prototypes of both converters have been implemented and evaluated in laboratory showing the superiority of the LLC resonant converter for almost the entire range of operation.

A constant power (CP)-constant voltage (CV) protocol for battery charging is implemented in a conventional boost converter with output filter (BOF) by imposing loss-free resistor (LFR) behavior during the CP phase. To compare on equal basis the performance of the new CP-CV technique with the classical constant current (CC)–CV protocol, the latter is also implemented in the same power stage. The CC phase in BOF is attained by imposing a G-gyrator of type II behavior to the converter. A versatile controller uses the same voltage regulation loop for both protocols during the CV phase and a slightly different loop for the CP and CC phases. The latter loop is based in both CP and CC phases on the sliding-mode control (SMC) of the input inductor current of BOF, which in steady-state is made proportional to the input voltage in the LFR case or to the output voltage in the gyrator implementation. To compensate for the slow variations of the battery voltage during the CC phase, a proportional-integral (PI) current regulator has been added in the gyrator realization. The comparison of the corresponding experimental results shows identical behavior in both approaches in the measured waveforms, component stress, efficiency and external temperature. The simplicity of the CP-CV implementation based on LFR allows the extension of the proposed protocol to other hard-switching converters.

Modern DC-DC power conversion represents an important challenge because connected loads are not purely resistive as it has been conventionally considered. Furthermore, the corresponding power converters perform functions which are not limited to regulate either a specific output voltage or output current. One of these new functions is the transfer of a regulated power to different types of loads, which emerge in the context of microgrids and electric vehicles, in which the sliding-mode control (SMC) is an important actor because of its versality, robustness and systematic design. This paper presents SMC of a boost converter operating as a loss-free resistor (LFR), which supplies a constant power to the parallel connection of three canonical elements, namely, a constant power load (CPL), a constant current load (CCL) and a dc voltage source with internal resistance. The studied load is defined as a generic nonlinear static load (GNSL). The subsequent analysis of the connection of the controlled power converter and the GNSL reveals the existence of a single equilibrium point, which is unconditionally stable. This feature is preserved when the GNSL is particularized in the single load cases of battery, current source or resistor, and in all cases of two-element and three-element load combinations. The exception is the supply of a single CPL, which results in an infinite number of equilibrium points with marginally stable behavior. Simulation and measurements in a 1 kW prototype are in perfect agreement with the theoretical predictions.

The unstable ON and OFF state trajectories of a boost converter supplying a constant power load (CPL) are appropriately combined using sliding-mode control (SMC) with a suitable surface to yield a stable trajectory that reaches a specified equilibrium point. Selecting the optimum surface involves a comparative analysis among potential candidates in terms of stability, disturbance rejection, conduction losses and inrush current. The performance as switching function of polynomials of degree zero (non-zero constant), degree one (affine function) and degree 2 (quadratic function) are analyzed in depth to conclude that the affine function leads to the best static and dynamic results. The theoretical predictions are verified by means of simulations and measurements in a prototype.

This paper presents a strategy to control the dynamic power sharing in a hybrid energy-storage system (HESS). The HESS is composed of a battery array of 24 V, a supercapacitor module of 16 V and 2 DC-DC converters allowing the connection of the system to an extra low voltage DC (ELVDC) bus of 48 V within a hybrid microgrid. The proposed strategy consists in a nested loop controller per converter enforcing charge and discharge regimes of the energy storage devices (ESD). The charging modes of the ESD are independent, whereas the discharging modes are coupled. Namely, the supercapacitor module injects power and regulates the ELVDC bus voltage, while the battery array injects the energy into the bus to maintain the levels imposed by the supercapacitor module until its contribution is cancelled. Furthermore, the battery array slowly injects the additional energy needed to restore the optimal supercapacitor voltage. Operation modes and transitions of the HESS control strategy are synthesized in a secondary control level. The entire proposal is validated through simulation results with scenario-based tests, where the modes of operations and the transient response to different loads and boundary conditions are verified.