arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1806.02543stat.MLcs.LG
收藏

太阳能发电预测的分组高斯过程

Grouped Gaussian Processes for Solar Power Prediction

University of New South Wales(新南威尔士大学);Data61(Data61研究所)

Astrid Dahl, Edwin V. Bonilla

摘要:

我们考虑多任务回归模型,其中的观测被假定为几个潜在的节点函数和权重函数的线性组合,这都是从高斯过程先验。开发可扩展的方法预测分布式太阳能和其他可再生能源发电的问题驱动下,我们提出了耦合先验组(节点或权重)的过程,利用功能之间的空间依赖性。我们估计多个分布式站点的太阳能发电预测模型和多个邻近气象站的地面风速。我们的研究结果表明,我们的方法相对于竞争对手的太阳能基准保持或提高了点预测精度,并在所有措施上提高了风预测基准模型。我们的方法始终占主导地位的等效模型没有耦合的先验,实现更快的收益预测精度。同时,我们的方法提供了更好的量化预测的不确定性。

英文摘要:

We consider multi-task regression models where the observations are assumed to be a linear combination of several latent node functions and weight functions, which are both drawn from Gaussian process priors. Driven by the problem of developing scalable methods for forecasting distributed solar and other renewable power generation, we propose coupled priors over groups of (node or weight) processes to exploit spatial dependence between functions. We estimate forecast models for solar power at multiple distributed sites and ground wind speed at multiple proximate weather stations. Our results show that our approach maintains or improves point-prediction accuracy relative to competing solar benchmarks and improves over wind forecast benchmark models on all measures. Our approach consistently dominates the equivalent model without coupled priors, achieving faster gains in forecast accuracy. At the same time our approach provides better quantification of predictive uncertainties.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑