Reactive power minimization of dual active bridge DC/DC converter with triple phase shift control using neural network.

Reactive power flow increases dual active bridge (DAB) converter RMS current leading to an increase in conduction losses especially in high power applications. This paper proposes a new optimized triple phase shift (TPS) switching algorithm that minimizes the total reactive power of the converter. The algorithm iteratively searches for TPS control variables that satisfy the desired active power flow while selecting the operating mode with minimum reactive power consumption. This is valid for the whole range of converter operation. The iterative algorithm is run offline for the entire active power range (-1pu to 1pu) and the resulting data is used to train an open loop artificial neural network controller to reduce computational time and memory allocation necessary to store the data generated. To validate the accuracy of the proposed controller, a 500-MW 300kV/100kV DAB model is simulated in Matlab/Simulink, as a potential application for DAB in DC grids