I. Introduction
In recent years, energy policies in many countries and districts are motivating end users to install photovoltaic (PV) systems for local energy consumptions [1], [2]. PV installations transform many users into prosumers who can both absorb energy from and feed energy to the main grid [3]. Since the integration of large-scale PV systems to the main grid may generate negative impact on the electric grid operations, most feed-in tariffs have been reduced to be much lower than the retail prices in order to encourage on-site consumption of renewable energy [2]. In addition, energy outputs of PV systems depending on the weather conditions are not generally aligned with load consumptions, which leads to the waste of solar energy and low local renewable energy efficiency [4], [5]. As a result, personal energy storages (PESs) such as Tesla [6] and Samsung [7] have been developed rapidly, and enable self-consumption of renewable energy by storing surplus energy for later consumptions. However, the investment cost of PESs is currently relatively high, and it is not economic for most porsumers to install their own PESs [2]. In addition, due to the preset capacities of PESs and PV systems with random outputs, these installations are hard to be fully utilized. An applicable way to promote local energy efficiency is to invest a community energy storage (CES) with relatively large capacity for local prosumers to share and utilize. Compared to PESs, the economic feasibility makes CES more applicable. Firstly, CES, usually with relatively larger capacity and lower average investment cost, is affordable for business investors. Secondly, the investor of the CES with larger capacity can make profits by providing storage service for community prosumers [5], [8]. Consequently, motivating energy sharing between CES and prosumers becomes the core to improve the community energy efficiency and economy.