Satisfactory model predictive control for a hybrid single-phase seven-level converter

摘要

Today, many hybrid multilevel converters with flying capacitors are being proposed. Besides the practicality of these converters in reducing the switching devices count and the cost, it is challenging to balance their capacitor voltage charge. The unbalanced capacitors restrain the converters' performance and lead to failure in delivering the required current. This paper proposes a model predictive control (MPC) for a seven-level converter based on an active neutral point clamped converter cascaded with an H-bridge. A conventional MPC uses a single cost function with two terms, one for current tracking and the other for capacitor balancing, which needs an accurate weighting factor to balance them. In this work, the suggested MPC adopts a satisfactory optimization technique. It evaluates the capacitor charge membership to a nominal voltage range to define the degree of freedom in which to optimize the current tracking problem. This transforms the relationship between the two cost terms into a more collaborative relationship. The proposed MPC improves the output current quality and balances the capacitor charge with the least number of computations. Experimental and simulation results have validated the controller's effectiveness.