Noise Tolerance Strategy Based on Virtual Capacitor for DC-DC Converters With Continuous Control Set Model Predictive Control

Noise can have a negative impact on the performance of dc-dc converters with continuous control set model predictive control (CCS-MPC). This issue is commonly encountered, making it difficult to effectively apply CCS-MPC in dc-dc converters. In this letter, we present a noise tolerance method for dc-dc converters with CCS-MPC that utilizes a virtual capacitor. This method offers a simple and effective solution to address the aforementioned problem while enhancing system robustness. To illustrate, we examine the noise generation mechanism and establish a predictive model using a dual-active-bridge converter as an example. Subsequently, we delve into the influence of noise on MPC and propose a noise tolerance method centered around the virtual capacitor. Importantly, this method does not incur additional costs, computational burden, or voltage/current ripples in the system, while preserving the inherent dynamic performance of MPC. Finally, we validate the effectiveness of the proposed method through experimental results using the TMS320F28377D as the core controller.