Modeling of power converters is usually done using cycle-average or piecewise-linear methods. Piecewise-linear methods operate by selecting the model corresponding to the current switching state. When the amount of possible states increase, the number of models increases as well. For model-predictive control (MPC), this results in an increasing computational effort, which limits the applicability of MPC to relatively simple models. This paper presents a Fourier-based modeling method to significantly reduce power-converter model complexity. Furthermore, a “voltage-balance” control algorithm is proposed using this method, which is applied to Active-Bridge type converters. The resulting closed-form algebraic solution yields a reduced circulating current and has considerably lower computational effort compared to piecewise-linear models (256 models vs. 1 model). An arbitrary amount of Active-Bridges can be used, which enables the use of scalable Active-Bridge converters. Results show a substantial reduction of the converter currents when the proposed method is compared to phase-shift control, which also significantly improves the EU efficiency, from 88.3% to 99.0% when applied to a QAB converter.