论文部分内容阅读
极值负荷的幅值与出现时刻是决定调度运行计划的重要依据，而其预测精度往往不尽如人意。实现概率化的预测，是规避极值负荷预测不准确所带来的风险的有效途径。以最高负荷为例，深入剖析了极值负荷的多峰特性，分析了幅值与出现时刻的统计规律，建立了日落时刻与晚高峰出现时间之间的回归模型；基于日间极值负荷增量的分类统计，预测下一天的各个子高峰幅值的概率分布（probabilistic density function，PDF），运用序列运算理论计算日最高负荷幅值的概率分布，最终根据全概率公式，实现了日最高负荷出现时刻的概率性预测方法。中国北方某城市的实际预测表明，所提出的概率化预测方法可以有效地解决极值负荷预测问题。“,”As the foundation of system daily scheduling and operations, current deterministic forecasting algorithm of the magnitude and timing of extreme load is not satisfactory. Probabilistic forecasting is an effective way to reduce the risk of inaccurate forecasting of extreme load. This paper took peak load as an example, analyzed the multi sub-peaks characteristic of load curve, studied the statistical features of the peak load magnitude and timing, established the regression model between peak load occurrence time and sunset time; based on sorted statistics of peak load daily increments by weeks and seasons, the paper forecasted the probabilistic density functions （PDF） of sub-peak load magnitudes, calculated the PDF of the peak load magnitude via sequence operation theories, forecasted the timing PDF employing total probability formula. Method proposed in this paper has been applied to a city in North China and the results prove the effectiveness of this method.