摘要
基于单项预测模型进行电网负荷预测已不能适应当前电网管理的要求。组合预测模型在很大程度上能够弥补单一预测方法的片面性,但在组合模型中固定负荷预测方法也存在预测不准确、可信度低等一系列问题。本文在电力负荷预测系统中引入动态组合的思想,通过自动筛选预测方法、动态配置权重,构建最优组合预测模型。实践证明,该组合预测方法比单个预测方法具有更高的预测准确性。
It has already failed to comply with the requirements on the network management to forecast the network load with the simple forecasting model.Combination forecasting model can largely make up for the one-sidedness of the simple forecast method,but fixing load forecasting method in combination model also exist inaccurate forecasts,low credibility and other problems.In this paper we applied the combined forecasting idea to power load forecasting system,built a combined optimal forecasting model through automatic sifting of the forecasting methods and weight dynamic configuration.Practice has proved that the combined forecasting method has higher forecast accuracy than a single forecast method.
作者
王战友
赵耀培
Wang Zhanyou;Zhao Yaopei(Shandong Management University,Jinan 250357,China)
出处
《信息化研究》
2018年第5期31-35,共5页
INFORMATIZATION RESEARCH
基金
山东省高校科研发展计划(J14LN17)
关键词
短期负荷预测
动态配置
组合预测模型
自动筛选
short-term load forecasting
auto filter
combination forecasting model
dynamic configuration