基于区间二型模糊逻辑的短期电力负荷预测  

Prediction of short-term power load using interval type-2 fuzzy logic

作　　者：王婧[1] 郑高[2] 吴拓[3] 薛宇[3] 

机构地区：[1]吉林大学珠海学院计算机科学与技术系,广东珠海519041 [2]西南交通大学电气工程学院,四川成都610031 [3]辽宁工程技术大学机械工程学院,辽宁阜新123000

出　　处：《兰州理工大学学报》2011年第3期108-113,共6页Journal of Lanzhou University of Technology

基　　金：广东省千百十工程资助项目;吉林大学珠海学院百人工程项目的资助

摘　　要：针对电力负荷的波动具有强随机性而难以准确预测的问题,引入区间二型模糊逻辑方法以减小预测误差.提出一个区间一型非单值二型模糊逻辑模型,用于一小时电力负荷的时间序列预测,并利用反向传播算法来调节模型输入及规则前、后件的参数.同时,构建一个基于反向传播算法的非单值一型模糊逻辑模型,将其预测结果作为检验性能的基准.仿真结果表明,区间一型非单值二型模糊逻辑模型提高了预测精度,性能优于相应的非单值一型模糊逻辑模型,验证了二型模糊逻辑方法能更有效地处理不确定性的观点.In order to solve the problem that power load is too stochastic to precisely forecast,an interval type-2 fuzzy logic method was introduced.An interval type-1 non-single-valued type-2 fuzzy logic model was presented for one-hour prediction of time series of power load.And the back-propagation（BP） algorithm was used for tuning model input and stipulating parameters of antecedent and consequent components.Meantime,a non-single-valued type-1 fuzzy logic model was set up on the basis of BP method and its prediction result was taken as a datum for performance checking.Simulation result showed that the interval type-2 fuzzy logic model exhibited a better performance than its type-1 counterpart due to its higher prediction accuracy,justifying the view point that the type-2 fuzzy logic method could deal with uncertainties in a more effective way.

关 键 词：电力负荷预测 时间序列预测 区间二型模糊逻辑 反向传播算法 

分 类 号：TM76[电气工程—电力系统及自动化] N94[自然科学总论—系统科学]