短期风速预测的相关方法及其应用研究

发布时间：2018-06-18 19:21

本文选题：风速预测 + 应用研究　；参考：《合肥工业大学学报(自然科学版)》2017年11期

【摘要】：文章对持续法、BP神经网络和支持向量机(support vector machine,SVM)3种方法在提前24h风速预测中的应用进行了研究和比较。为了消除季节对预测结果的影响,针对某年12个月份分别建立预测模型,结果表明:在大多数情况下,BP神经网络和SVM算法的预测结果要优于持续法,并且SVM算法优于BP神经网络;但也有持续法优于BP神经网络和SVM算法及BP算法优于SVM算法的情况。因此不能绝对说某种算法优于另一种算法,应该根据具体情况来进行分析判断,或者通过组合预测来提高预测精度。
[Abstract]:In this paper, the application of continuous BP neural network and support vector machine SVM in 24 hours' wind speed prediction is studied and compared. In order to eliminate the influence of seasons on the prediction results, the prediction models are established for 12 months of a given year. The results show that in most cases, the prediction results of BP neural network and SVM algorithm are better than that of the persistence method. The SVM algorithm is superior to the BP neural network, but there are some cases where the persistence method is superior to the BP neural network and SVM algorithm and the BP algorithm is superior to the SVM algorithm. Therefore, it can not be absolutely said that one algorithm is superior to the other, and it should be analyzed and judged according to the specific situation, or the prediction accuracy should be improved by combination prediction.
【作者单位】：国网安徽电力公司检修公司;合肥工业大学教育部光伏系统工程研究中心;
【基金】：国家重点基础研究发展计划(973计划)资助项目(2009CB219708) 国家自然科学基金面上资助项目(51077033) 高等学校博士学科点专项科研基金资助项目(201301111110005) 广东省引进创新团队资助项目(2011N015)
【分类号】：TM614

【相似文献】