%0 Journal Article
%T 基于支持向量分位数回归神经网络的区域短期风速预测<br>A Regional Short-Term Wind Speed Prediction Based On Support Vector-Quantile Regression Neural Network
%A 生菡
%A 秦喜文
%J Advances in Applied Mathematics
%P 3325-3336
%@ 2324-8009
%D 2021
%I Hans Publishing
%R 10.12677/AAM.2021.1010349
%X 由于风速的间歇性导致了风电系统输出的波动性，可靠的风速预测可以有效提高风电装置的稳定性。在分位数回归的框架下，提出了支持向量–分位数回归神经网络(SV-QRNN)模型。首先，采用粒子群优化算法对支持向量机参数进行优化，得到支持向量。然后采用分位数回归神经网络算法推导出风速的条件分位数，以及不同置信度下的预测区间。实验结果表明，SV-QRNN模型能够较好地平衡预测性能和时间效率。同时还可以补贴发电调度决策，并允许电力市场充分发挥作用。<br />
Due to the intermittent nature of wind speed, reliable wind speed prediction improves the volatility of wind system output due to wind speed instability. This paper proposes a support vector-qu- antile regression neural network (SV-QRNN) model under the framework of quantile regression. To begin with, particle swarm optimization algorithm is employed to optimize the parameters of support vector machine to obtain the support vector. Then the quantile regression neural network algorithm is adopted to derive the conditional quantile of wind speed, and the prediction interval under different confidence levels is obtained. The experimental results illustrate that the SV-QRNN model can better balance the prediction performance and time efficiency by comparing several other competitive models based on interval prediction. It can also subsidize generation scheduling decisions and allow the full performance of the electricity market.
%K 分位数回归神经网络，支持向量，风速，粒子群优化算法<br>Quantile Regression Neural Network
%K Support Vector
%K Wind Speed
%K Particle Swarm Optimization Algorithm
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=45813