Abstract:The randomness, volatility and anti-peaking characteristic from distributed renewable energy generation rise great challenges for the safe and economic operation of the distribution networks (DNs). To address this problem, this paper proposes a novel multi-energy-storage energy management system (EMS) to co-optimize the electricity-driven mobile energy storage (MES) and inverter air-conditioning (AC)-based thermal energy storage (TES). To facilitate the energy management of the DN, the MES that considers the delay factors and the TES that regulates reactive power have been developed into a unified analytic model capable of charging and discharging. In addition, considering the impact of the forecasting uncertainties and the risk-aversion of the dispatcher, a novel robust optimization method is proposed to obtain more accurate "worst scenario". The dispatching model is then converted into a mixed integer second-order cone programming problem (MI-SOCP) and a mixed integer linear programming problem (MILP), and linearized techniques and an iteration method are used to efficiently solve these problems. Simulation studies on a 41-node DN in Ontario indicate that the operational cost and power loss of the DN can be reduced by no less than 1% and 8% using the proposed EMS, respectively, while a safer voltage level with a voltage deviation of 5% can be obtained. The results confirm the effectiveness of the MES and TES for peak shaving, valley filling and voltage supporting.

Multi-energy-storage Energy Management with the Robust Method for Distribution Networks