With the increasing penetration of renewable energy sources and active loads into the electrical power system, the limitations of traditional low-frequency power grids are becoming apparent. The solid-state transformer (SST), with its multiple conversion stages offering various power ports, is an innovative solution to these challenges. However, the high cost of the SST, due to its heavy reliance on power electronic components, makes the complete replacement of traditional grids an economically nonviable option. Therefore, this paper proposes a parallel connection of the SST to an existing traditional power grid in order to divide the transmitted power, thus reducing the investment costs. Furthermore, a hybrid topology based on modular multilevel converter (MMC) and shunt active power filter (SAPF) is proposed in this paper in order to compensate the harmonic currents caused by non-linear loads while maintaining steady energy delivery. Additionally, an artificial neural network (ANN) controller is adopted for the MMC output DC voltage control to ensure a regular DC-link voltage for both the SAPF and the DC-to-DC converter of the SST. The proposed topology is implemented and evaluated in MATLAB/Simulink software, and various simulation results are presented.
Read full abstract