STATE OF CHARGE BALANCING METHODS FOR DISTRIBUTED BATTERY ENERGY UNITS IN DC MICROGRID

Abstract

Recently, the expansion of renewable energy sources (RESs) in total electricity generation structure along with the development of power electronics technology and the increase of various types of DC loads lead to the growth of the DC microgrid as an appropriate model for future energy systems, especially for the locations that the main grid cannot meet. The advantages of DC microgrid includes less complicated control system, no issues relating to reactive power and frequency synchronization. In DC microgrid, RESs generate the power fluctuation since their intermittent nature. In order to guarantee the stable operation of DC microgrid, the power fluctuation caused by RESs is usually mitigated by using distributed battery energy units (BEUs). Due to the voltage drop induced by line resistance in DC microgrid, the BEUs operate with different charging or discharging power, which leads to the imbalanced state of charge (SoC) among BEUs, and the SoC imbalance can cause a certain BEU become over-discharged or over-charged. In this thesis, control strategies are developed to achieve the SoC balancing among BEUs in DC microgrid along with the accurate power sharing and voltage compensation. Firstly, an enhanced control method is proposed to balance SoC and achieve the accurate power sharing of distributed BEUs with the same capacities in a DC microgrid. In this method, the virtual power rating concept is presented, and accurate power sharing is obtained in accordance with SoC values and virtual power ratings, despite of different line resistance in the microgrid. In addition, voltage compensation is considered to maintain the DC microgrid voltage within a desired level. Secondly, a SoC balancing method based on the adaptive virtual power rating method is presented. The virtual power rating is flexibly regulated by SoC in order to develop a modified droop control method for distributed BEUs. The SoC balancing and accurate power sharing among BEUs are satisfied regardless of the line resistance difference. In addition, the virtual power rating is kept within the actual power rating of BEU to avoid overload, and a simple voltage restoration method without PI controller is presented to keep the grid voltage at a desired value. Thirdly, an adaptive droop control is introduced to balance SoC among distributed BEUs with different capacities in a DC microgrid. In this method, the droop coefficient of BEU is regulated not only by its SoC but also its capacity, and the corresponding virtual power rating is also determined to realize the accurate power sharing along with the SoC balancing among BEUs. Moreover, a voltage restoration method is introduced to maintain the microgrid voltage level within an allowable range. Finally, all proposed methods are implemented as the distributed control strategies by using the low bandwidth communication network, and they are verified by proper simulations and experiments.