Adaptive Control of Bidirectional DC-DC Converters for EV Applications

Abstract

The increasing trend of electric vehicles (EVs) activity has increased the demand of effective, reliable, and dynamically variable power management structures, and one such aspect is the bidirectional DC-DC converters (BDC) which are vital in transferring energy between the traction package battery and the DC-link in the vehicle. Conventional fixed-gain control techniques (e.g. PI and SMC), typically cannot keep up the performance requirements under variable loads, wide swings in battery state-of-charge (SoC), and quickly changing operating states (e.g. when performing regenerative braking, and during a transient acceleration). In this paper, an adaptive control approach, which is Model Reference Adaptive Control (MRAC), enriched using the Lyapunov stability methodology is proposed, in particular in relation to real-time regulation of the voltage and current flow in both buck (regenerative) and boost (motoring) operation. The implementation is based on the model reduction of current-controlled converter system through averaged state-space representation and design of an adaptive controller that could modify its parameters based on external perturbations and internal parameter changes. MRAC structure will consider the system states to behave according to the given reference model and offer global asymptotic stability of the state by adoption of a Lyapunov development based adaptation law. The control scheme is designed and the validation of the same is performed by extensive simulation studies with MATLAB/Simulink and the test scenarios were based on standard urban driving cycles The various performance data, settled time, overshoot, and current ripple, are also compared with the conventional control methods and it is seen that the new concept has better robustness, quicker transient response and better voltage regulation. Also, hardware-in-the-loop (HIL) testing takes place on a real-time simulation platform and an embedded controller to confirm feasibility in real practice. The performance indicates that the adaptive controller can ensure high tracking accuracy against mode-switching and in nonlinear disturbance environments, which can ensure energy minimization and increase the life of the batteries in EV. In (research) terms, it introduces a scalable solution that can be implemented in real-time to the next-generation EV power converters and guides future research that would incorporate predictive and data-driven control layers into predictive and adaptive EV energy management systems.