Modified Model Predictive Control for Power Balancing in Multilevel Inverter for Battery Energy Storage Applications

Battery energy storage systems (BESS) will play a critical role in the modern power system (MPS) with high penetration of renewable energy resources. In MPS energy paradigm, it is required to deploy and integrate BESS at different voltage level from low voltage distribution system to medium voltage levels. This requires high number of battery cells in series to reach to desired voltage level for grid integration which is challenging from BESS technology point of view and unequal state of health of batteries over time. On the other hand, cascaded multilevel inverters (CMI) provides a unique opportunity for minimizing the direct series connection of battery cells for medium voltage grid integration of BESS. However, conventional CMI’s linear and non-linear control schemes suffers from unequal power drawn from battery cells at each h-bridge. This paper proposes a modified predictive control scheme that leverages the inherent characteristics of finite-set model predictive control (MPC) and balance power drawn from the battery cells. Conventional MPC is based on a cost function optimization that select the first optimal switching state, this approach will result in unequal power drawn form battery cells. The proposed approach includes a constraint optimization that mitigates the unequal power drawn from battery cells. The theoretical analysis are validated via several case studies, the results demonstrate promising performance of proposed modified MPC for CMI interfaced BESS.