MODEL PREDICTIVE CONTROL OF A BIDIRECTIONAL DUAL ACTIVE BRIDGE CONVERTER FOR ELECTRIC VEHICLE CHARGING

Abstract

This thesis presents the complete design, modeling, control, and MAT LAB/ Simulink-based implementation of an 11 kW bidirectional Dual Active Bridge DC-DC converter intended for electric vehicle charging applications under both Grid to-Vehicle (G2V) mode of operation. The converter is designed for a 400 V input DC-link, a 350–400 V battery voltage range, a switching frequency of 100 kHz, a leakage inductance of 50 µH, and an output capacitance of 500 µF. A Lithium-ion bat tery model of 60 Ah capacity has been considered for the analysis. Single-Phase-Shift (SPS) modulation is employed to control the magnitude and direction of the transferred power. A finite-horizon Model Predictive Controller (MPC) has been developed and implemented within a MATLAB Function block. The controller employs a discrete-time linearised state-space prediction model of the DAB converter, a quadratic cost function, and explicit constraints on phase-shift angle, rate of change, and inductor current. The optimization problem is solved in a receding-horizon fashion with a Quadratic Programming (QP) solver. An RC-based compensation network is proposed to enhance the closed-loop stability and dynamic response. Lead, lag, pole-zero cancellation, feedforward, pole splitting, digital and current-mode compensation methods are also discussed. The whole system has been simulated in MATLAB/Simulink using Sim scape Electrical and Specialized Power Systems libraries. .