摘要
由于风速的随机性大,预测的准确度不高,针对这种现象,基于支持向量机理论建立了风速预测模型,同时针对支持向量机参数的选取尚无有效的方法,尝试应用蚁群算法来优化参数的选取。以某风场连续5 d的实测风速为研究对象,选取前4 d的实测风速(采样间隔30 min),应用风速预测模型对第5天的48个风速值进行预测,其预测的平均绝对百分比误差仅为9.77%,预测效果比较理想。验证了应用蚁群优化算法理论与支持向量机理论进行风速预测的可行性,可为风速的长期预测、风力发电功率预测和风电场规划选址等提供理论指导。
As the great randomness and low forecasting accuracy of wind speed,a wind speed forecasting model is established based on support vector machine theory,and according to the lack of an effective method for its parameter selection,the ant colony algorithm is used to optimize the selection.Taking the measured wind speed of continuous 5days in a wind farm as an example,we forecast 48 wind speed values of the fifth day using the established model,based on the measured samples of the first four days,and the absolute percentage error is merely 9.77%,which is satisfying.Therefore,the feasibility of wind speed forecasting by support vector machine and ant colony optimization is verified,which could provide a theoretical guideline for long-term forecasting of wind speed,wind power generation forecast and site selection of a wind power farm.
出处
《人民长江》
北大核心
2011年第4期95-97,101,共4页
Yangtze River
关键词
支持向量机
风速预测
蚁群优化算法
风电场
风力发电
新能源
support vector machine
wind speed forecast
ant colony optimization algorithm
wind power farm
wind power generation
new energy
