摘要
采用概率分布拟合和基于遗传算法的BP神经网络的预测建模对风电功率波动特性进行定量分析。首先,针对风电功率实测数据建立了概率分布拟合模型并分析了拟合结果;其次,建立了基于不同时间间隔历史实测数据的BP神经网络预测模型,数据检验表明该模型对于峰值有很理想的预测精度且整体精度较高。
The fluctuation of wind power is analyzed by probability distribution fitting and forecasting model of BP neural networkbased to quantitative. Firstly, the distribution fitting probability model is established according to the measured data of wind power; secondly, BP neural network prediction model for different time intervals of history data is established; the existing data test show that the model for the peak has a very good prediction and the overall prediction accuracy is higher. The obtained results have certain guiding significance to create balance mechanism of wind power effectively.
出处
《湖南文理学院学报（自然科学版）》
CAS
2015年第2期39-42,共4页
Journal of Hunan University of Arts and Science(Science and Technology)
基金
国家自科基金(11101053
71471020)
国家级大学生创新训练项目(201410536012)
长沙理工大学教改项目(JG 1242)
关键词
概率分布拟合
遗传算法
BP神经网络
probability distribution fitting
genetic algorithm
BP neural network
