Zhan Cun, Wei Renjuan, Zhao Lu, Chen Shijun, Shen Chunying
Faculty of Electric Power Engineering, Kunming University of Science and Technology, Kunming, 650500, China.
Sichuan Water Conservancy Vocational College, Chengdu, 611200, China.
Sci Rep. 2025 Feb 7;15(1):4648. doi: 10.1038/s41598-024-83798-2.
Increasing drought events have threaten electricity supply security in the predominantly hydropower-based Sichuan Province. Wind power has the potential to complement hydropower, yet its complex fluctuations required a systematic assessment. Accordingly, we evaluated maximum likelihood estimation and three goodness-of-fit tests to identify the optimal distribution model of daily wind speed records during 1961-2017 across 156 weather stations in Sichuan Province among six commonly used probability density distributions. The study further analyzed the spatiotemporal features of persistence and multifractality in wind speed records across various landform types using multifractal detrended fluctuation analysis. The principal outcomes of our study indicated that the generalized extreme value distribution served as the optimal model for fitting wind speeds in Sichuan Province, outperforming the commonly used Weibull distribution. Persistence was evident in all wind speed series as the Hurst index exceeds 0.5, with the strongest persistence in mountainous areas and the weakest in plains. Multifractality was confirmed by the non-linear dependencies of the Generalized Hurst Exponent [h(q)] and mass exponent [τ(q)] on q, as well as by the multifractal spectrum widths exceeding 0.05. Among landform types, plains exhibited the strongest multifractality, followed by plateaus, with mountains showing the weakest multifractality. Long-range correlations were identified as the primarily caused of multifractality, as indicated by narrower multifractal spectrum widths in both shuffled and surrogate series, and stronger narrowness in the shuffled series. The multifractal spectrum width of the mountain shuffle series, which slightly exceeded 0.05, further highlighted the determinative influence of long-range correlations. Considering these findings, the southwestern mountainous region emerges as the optimal area for wind farm development, given its stability (persistence) and moderate fluctuation complexity (multifractality), crucial for effective wind resource utilization in hydropower-dominated settings. Our study provides a novel approach to assessing wind resources and offers guidance for wind farm placement in complex terrain regions, supporting sustainable energy diversification in Sichuan Province.
日益增多的干旱事件已经威胁到了以水电为主的四川省的电力供应安全。风能有潜力补充水电，但它复杂的波动需要进行系统评估。因此，我们评估了最大似然估计和三种拟合优度检验，以便在六种常用概率密度分布中确定1961 - 2017年期间四川省156个气象站日风速记录的最优分布模型。该研究还使用多重分形去趋势波动分析，进一步分析了不同地形类型风速记录中持续性和多重分形的时空特征。我们研究的主要结果表明，广义极值分布是拟合四川省风速的最优模型，优于常用的威布尔分布。由于赫斯特指数超过0.5，所有风速序列都存在持续性，山区持续性最强，平原地区最弱。广义赫斯特指数[h(q)]和质量指数[τ(q)]对q的非线性依赖性以及多重分形谱宽度超过0.05，证实了多重分形性。在地形类型中，平原地区的多重分形性最强，其次是高原，山区的多重分形性最弱。重排序列和替代序列的多重分形谱宽度都变窄，且重排序列更窄，这表明长程相关性是多重分形性的主要原因。山区重排序列的多重分形谱宽度略超过0.05，进一步突出了长程相关性的决定性影响。考虑到这些发现，西南山区因其稳定性（持续性）和适度的波动复杂性（多重分形性），成为风电场开发的最佳区域，这对于水电主导地区有效利用风能资源至关重要。我们的研究提供了一种评估风能资源的新方法，并为复杂地形地区风电场的选址提供了指导，支持四川省可持续的能源多样化。
