论文部分内容阅读
采用希尔伯特黄变换和Mann-Kendall非参数检验法对对澜沧江流域大朝山电站1961~2008年入库径流周期及趋势变化规律进行研究,根据分析结果建立BP神经网络希尔伯特黄变换耦合模型预测了未来5 a大朝山电站入库径流。结果表明,大朝山电站入库流量序列主要存在3、7、16 a近似周期成分,1961~2000年入库流量呈减少趋势,在进入21世纪后径流有所增加,未来径流量将继续保持递减趋势,2014年将达到最小值,之后径流有所回升。
The Hilbert Huang transform and Mann-Kendall nonparametric test were used to study the variation regularities of the incoming flow and the trend of the Dachaoshan Hydropower Station from 1961 to 2008 in the Lancang River Basin. Based on the analysis results, the BP neural network Hilbert transform The coupled model predicts the runoff of the Dachaoshan Hydropower Station in the next 5 years. The results show that there are mainly 3,7,16 a approximate periodic components in the incoming flow sequence of the Dachaoshan Hydropower Station and the decreasing trend of the stock flow in 1961 to 2000. The runoff increases after the 21st century and the runoff will continue to decrease in the future The trend will reach a minimum in 2014, after which the runoff will pick up.