Data-driven Economic Control of Battery Energy Storage System Considering Battery Degradation

Abstract

Battery energy storage system (BESS) provides a potential solution to mitigate real-time power imbalance by participating in frequency control. However, the battery aging resulted from intensive charge-discharge cycles will inevitably lead to batter degradation, which could potentially incur high control cost. It is therefore important to design an optimal control method for battery energy storage systems (BESS) which can achieve power balance and consider tradeoff of battery aging simultaneously. To achieve this purpose, this paper employs a data-driven approach based on deep reinforcement learning to design an economically optimal controller for BESS. An actor-critic model is proposed for optimizing the BESS controller performance. A cost model based on battery cycle aging cost, unscheduled interchange price and generation cost is employed to estimate the operational cost of BESS in power systems. The effectiveness of the proposed optimal BESS control method is verified in a three-area power system.