System identification and adaptive control of a DC-DC converter using a current balancing ON/OFF control technique for optimal transient performance

This paper presents a novel adaptive control technique for the output voltage regulation of a dc-dc buck converter application. In particular, it focuses on the impact of unexpected system parameter variation on the static and dynamic performance of the system. We apply a Fast Affine Projection (FAP) technique for the purposes of parameter estimation. This technique has previously been shown by the authors to offer excellent performance as a system identification algorithm for power electronic systems. The FAP algorithm is computationally efficient, exhibits fast parameter estimation convergence, and is highly accurate. The estimated parameters are used to develop a novel Current Balancing ON/OFF (CBOO) switching strategy during transient periods of operation. Here, constant “ON” and constant “OFF” periods are calculated, based on the specific parameter estimation output to achieve optimal control performance and significantly improve the dynamic response of the dc-dc converter. Simulation results demonstrate that, in comparison to classic PI control schemes, the proposed method exhibits superior dynamic performance.