Kalman Filter-Based Model-Free Predictive Control of Classical DC–DC Power Converters

Conventional model predictive control (MPC) of power converters has been widely found in many power electronics and motor drive applications. The performance of MPC strongly depends on the precision of the converter’s physical parameters, and a mismatch of them produces a control degradation, which leads to MPC suboptimal operation. Ensuring a precise estimation of the converter’s parameters is difficult because they continuously change during the operation process due to their operating point and aging. Recently, model-free predictive control (MF-PC) has been used in motor drives and power electronics converters, especially inverters and rectifiers, to deal with the predictive control method’s dependency model. However, MF-PC proposed for dc–dc converters is an open innovation scientific field. This article proposes an MF-PC designed for second-order dc–dc converters, such as the boost, buck, buck–boost, and noninverting buck–boost converters. The presented approach uses a Kalman filter to estimate the positive and negative inductor current slopes with high accuracy and a low computational cost. The experimental results show that the proposed method is robust against parameter and model changes compared to conventional model-based solutions.