An interactive operation management of a micro-grid with multiple distributed generations using multi-objective uniform water cycle algorithm
An interactive operation management of a micro-grid with multiple distributed generations using multi-objective uniform water cycle algorithm
复制标题
DOI：
10.1016/j.energy.2016.03.048
复制
发表时间：
2016-07
期刊：
影响因子：
9
通讯作者：
A. Deihimi;B. Zahed;R. Iravani
中科院分区：
文献类型：
--
作者：
A. Deihimi;B. Zahed;R. Iravani
分布式电源(DG)在靠近负荷的情况下运行，从而产生了微电网(MG)的概念，以提高能源供应的可靠性和质量。MG作为一群用户和DG可以独立运行和并网运行，经常需要ESS(储能系统)来处理发电余缺。随着可再生能源和能源消耗的变化，以及经济和环境问题，MG需要一个有效的OM(操作管理)来对DG、ESS和上游宏观电网的交换路径进行短期能源调度。本文提出了以运行成本和排放量为目标的多目标均匀水循环算法(MOUWCA)。这个问题被用来寻找24个POF(帕累托最优前沿)，对应于一天中的24个小时(不像以前的研究每天给出一个POF)，以提供更多的灵活性来选择每小时折衷的解决方案。通过交互过程，根据排放ESS所需的小时顺序，在一天内平衡ESS的充电/排放。在一些基准问题上对MOUWCA进行了测试，并与非支配GA-II(NSGA-II)、多目标粒子群算法(MOPSO)和正常约束算法(NCA)进行了比较，以验证其有效性。然后将MOUWCA应用于一个典型的MG，通过与其他以前使用的算法的比较，证实了该算法的优越性。
Accommodation of DGs (distributed generations) close to loads has led to the concept of MG (micro-grid) for better reliability and quality of energy supply. MG as a clump of consumers and DGs can operate in stand-alone and grid-connected modes, and often needs ESS (energy storage system) to handle generation surplus/shortage. Variations of renewable sources and consumptions along with economical and environmental issues necessitate an efficient OM (operation management) of MG for short-term scheduling of energy outputs of DGs, ESS and exchange route to upstreammacro-grid. This paper presents MOUWCA (multi-objective uniform water cycle algorithm) for optimal OM of MG considering operation cost and emission as objectives. The problem is casted to find 24 POFs (pareto-optimal fronts) corresponding to 24 h of the day (unlike previous studies giving one POF per day) to provide more flexibility for selecting hourly compromise solutions. Through an interactive process, charging/discharging of ESS is balanced over a day based on the desired order of hours for discharging ESS. MOUWCA is examined on some benchmark problems and compared with NSGA-II (non-dominated GA-II), MOPSO (multi-objective particle swarm optimization) and NCA (normal constraint algorithm) to verify its effectiveness. MOUWCA is then applied to a typical MG where its superiority is confirmed in comparison to other previously used algorithms.