摘要
基于电量数据内在规律及与外界环境变量关系的深度挖掘,定义了占季比指标,提出了最小二乘支持向量机算法与基于占季比预测相结合的电量预测混合算法;根据国内A省历史分月电量进行实际算例分析,通过Elman神经网络、BP神经网络预测方法与文中所提方法的预测误差对比,验证了所提算法的预测精度,证明所提算法对预测精度有较大提高。
Basing on the inherent law of electric data and the relationship of external environment; theelectricity month-season ratio （EMSR） is defined. Proposing the improved power prediction algorithm that combines least squares support vector machine algorithm and the quarterly EMSR forecast algorithm. The monthly power data of province A is used to do the verification. By comparing the forecast error of Elman neural network, BP neural network with the forecast error of proposed algorithm. The accuracy of the proposed algorithm is verified, which shows that the accuracy has been improved.
出处
《电网与清洁能源》
北大核心
2017年第3期71-76,共6页
Power System and Clean Energy
基金
国家自然科学基金项目(51507141)~~
关键词
电量预测
ELMAN神经网络
最小二乘支持向量机
占季比
least squares support vector machine
electricity consumptionforecasting
month-season ratio
Elman neural network
