基于模糊概率的微电网短期负荷预测方法  

Short Term Load Forecasting Method of Microgrid Based on Fuzzy Probability

作　　者：黄锦增 段炼 邓祺 HUANG Jinzeng;DUAN Lian;DENG Qi(Guangzhou Power Supply Bureau of Guangdong Power Grid Co.,Ltd.,Guangzhou 510630,China)

机构地区：[1]广东电网有限责任公司广州供电局,广东广州510630

出　　处：《微型电脑应用》2023年第4期187-190,共4页Microcomputer Applications

摘　　要：在采集微电网历史负荷数据时,常规方法得到的负荷序列异常值较多,导致负荷预测值与实际值的均方误差、绝对百分比误差较高,因此提出基于模糊概率的微电网短期负荷预测方法。获取微电网历史负荷数据,通过平滑处理,修正历史数据异常负荷点,选取天气、日类型作为负荷影响因素,采用模糊概率,计算影响因素最大概率量化值,将负荷值和量化值作为输入向量,输入至径向基神经网络中,迭代训练后输出短期负荷预测值。选取工业场景和居民场景微电网,设置对比实验,结果表明,设计方法降低了负荷预测值与实际值的均方误差、绝对百分比误差。When the conventional method is used to collect the historical load data of the microgrid,there are many abnormal values of the load sequence,resulting in a high mean square error and absolute percentage error between the predicted value of the load and the actual value.Therefore,a short-term load forecasting method for the microgrid based on fuzzy probability is proposed.It obtains the historical load data of the microgrid,modifies the abnormal load points of the historical data through smoothing processing,selects the weather and day types as the load influencing factors,uses the fuzzy probability to calculate the maximum probability quantitative value of the influencing factor,and uses the load value and the quantitative value as the input vector.These data are input to the radial basis function neural network,and the network outputs the short-term load prediction value after iterative training.The industrial scene and residential scene microgrid are considered,comparative experiments are carried out,the results show that the design method reduces the mean square error and absolute percentage error between the load forecast value and the actual value.

关 键 词：模糊概率 微电网 负荷预测 神经网络 

分 类 号：TM734[电气工程—电力系统及自动化]