基于SCM-ANFIS负荷预测的储能电站调峰控制策略  

PEAK REGULATION CONTROL STRATEGY OF ENERGY STORAGE POWER STATION BASED ON SCM-ANFIS LOAD FORECAST

作　　者：王晓东[1] 苗宜之 卢奭瑄 刘颖明[1] Wang Xiaodong;Miao Yizhi;Lu Shixuan;Liu Yingming(School of Electrical Engineering, Shenyang University of Technology, Shenyang 110870, China;School of Chemical Process Automation, Shenyang University of Technology, Liaoyang 111003, China)

机构地区：[1]沈阳工业大学电气工程学院,沈阳110870 [2]沈阳工业大学化工过程自动化学院,辽阳111003

出　　处：《太阳能学报》2018年第6期1651-1657,共7页Acta Energiae Solaris Sinica

基　　金：国家自然科学基金(51677121);辽宁省自然科学基金指导计划(201602549);辽宁省教育厅资助项目(LGD2016031)

摘　　要：针对电池储能调峰控制中预测负荷存在误差较大、峰谷识别不准等问题,提出一种基于基于减法聚类和自适应网络模糊推理（SCM-ANFIS）电网负荷预测和调峰目标动态规划相结合的储能电站调峰控制策略。负荷预测中采用减法聚类减少模糊规则数目,并通过混合学习算法训练神经网络参数,从而减小预测计算量,提高预测精度。调峰过程中基于负荷预测信息通过引入分阶段滚动优化,在储能系统容量约束下实现调峰效果最优。基于某区域实际电网负荷数据的算例结果验证预测算法和控制算法的可行性和有效性。Aiming at the problems of large errors in the predicted load and inaccurate identification of peaks and valleysin battery energy storage peak regulation control,the peak regulation control strategy of energy storage power stationcombining power grid load forecasting and peak regulation target dynamic planning was proposed based on subtractiveclustering method and adaptive network fuzzy inference system（SCM-ANFIS）. The subtractive clustering method wasused to reduce the number of fuzzy rules in load forecasting,and neural network parameters were trained by hybridlearning algorithm,thereby reducing the amount of prediction calculation and improving the prediction accuracy. In theprocess of peak regulation,based on the load forecasting information,by introducing the phased rolling optimization,thepeak regulation effect is optimal under the constraint of the capacity of the energy storage system. Based on the results ofthe actual grid load data in a certain area,the feasibility and effectiveness of the prediction algorithm and controlalgorithm were verified.

关 键 词：储能 负荷预测 模糊推理 削峰填谷 多阶段目标动态规划 

分 类 号：TM919[电气工程—电力电子与电力传动]