摘要
电力负荷具有较大的波动性,且负荷数据存在噪声、缺失或异常值等问题,影响概率区间预测结果的准确性。为此,提出一种优化的电力动态负荷数据概率区间预测方法,并对该方法展开验证分析。根据当前预测需求及标准的变化,对电力动态负荷数据预处理,计算动态负荷数据概率感知区间合适值,压缩概率感知区间范围。采用多层级的方式,细化负荷数据概率区间预测范围。设计电力动态负荷数据概率区间预测流程,采用多目标PF-Elman集成处理,实现概率区间预测。测试结果表明,此次所设计的动态负荷数据概率区间预测方法最终得出的APFE值均被较好地控制在4以下,当前所设计的预测方法更加灵活、高效,预测的效率更佳,预测结果的区间覆盖率都在95%以上,具有实际的应用价值。
The power load has large volatility,and the load data has problems such as noise,missing or outlier,which affects the accuracy of the probability interval prediction results.Therefore,an optimized method for probability interval prediction of power dynamic load data is proposed,and verified and analyzed.According to the changes of the current forecast demand and standards,the power dynamic load data is preprocessed,the appropriate value of the probability perception interval of the dynamic load data is calculated,and the range of the probability perception interval is compressed.A multi-level approach was used to refine the prediction range of the probability interval of the load data.The probability interval prediction process of power dynamic load data is designed,and the multi-objective PF-Elman integrated processing is used to achieve the probability interval prediction.The test results show that the APFE value of the dynamic load data probability interval prediction method is well controlled below 4.The current designed prediction method is more flexible and efficient,the prediction efficiency is better,and the interval coverage of the prediction results is more than 95%,which has practical application value.
作者
许陈德
XU Chende(Guangzhou Power Supply Bureau,Guangzhou 510000,China)
出处
《传感器世界》
2024年第1期23-28,共6页
Sensor World
关键词
电力动态负荷
数据概率
区间预测
预测方法
负荷控制
power dynamic load
data probability
interval prediction
prediction method
load control