Papers09:52Chinese content高效包装器特征选择提升风电和太阳能预测实证研究这篇论文提出了CSFS方法,能在不牺牲预测精度的情况下减少21%的计算成本,适合做风电和太阳能预测的朋友参考。#CSFS#feature selection#wind power forecasting#solar power forecastingaarXiv cs.AI@Daniel Grillmeyer et al. (6)SourceRead laterReadWorth followingUsefulFollow CSFS
Papers11:05Chinese content深度学习模型在NWP误差下光伏功率预测的鲁棒性分析这篇论文用具体模型和实验告诉你,哪些深度模型在光伏功率预测里更扛NWP天气误差,不是空谈。#PatchTST#GRU#NWP#solar power forecastingaarXiv cs.LG@Dandan Chen et al. (3)SourceRead laterReadWorth followingUsefulFollow PatchTST
PapersSelected10:40Chinese content基于物理信息合成历史的时间序列基础模型用于光伏冷启动预测光伏运维团队终于有了冷启动预测的实用方案——无需历史数据即可用基础模型实现高精度预测,做新能源发电预测的开发者可以直接参考TabPFN-TS和Chronos-2的表现。#time series foundation models#solar power forecasting#cold start#zero-shotaarXiv cs.LG@Lorenzo Longarini et al. (5)SourceRead laterReadWorth followingUsefulFollow time series foundation models