An Overview of Data Preprocessing for Short-Term Wind Power Forecasting
An Overview of Data Preprocessing for Short-Term Wind Power Forecasting
复制标题
短期风电预测数据预处理概述
DOI：
10.1109/icasi52993.2021.9568453
复制
发表时间：
2021
期刊：
影响因子：
--
通讯作者：
Q. Phan
中科院分区：
文献类型：
--
作者：
Quoc;Yuan;Q. Phan
风力发电因其环境效益和经济效益在电网中发挥着越来越重要的作用。然而，将风力发电纳入电力系统的主要挑战包括可变性和不确定性。准确的预测可以降低运行成本，提高电力系统的稳定性。风电预测包括数据采集、数据预处理、模型构建与训练、误差计算等多个步骤。其中，数据预处理在风电预测过程中起着重要的作用，因为预测模型的输入对数据质量很敏感。因此，本文对风力数据处理方法进行了综述。这些方法的目的是对数值天气预报（NWP）风速和实测风电数据进行预处理并提取适合的特征。最后，通过实例分析说明了预处理步骤在风电预测中的重要性。
Wind power generation takes on an increasingly vital role in the power grid due to its environmental and economic benefits. However, the primary challenges that are related to the integration of wind power into power systems include variability, uncertainty. An accurate forecasting reduces operating costs and enhances power system stability. Wind power forecasting include many steps, including data collection, data preprocessing, the construction and training for models, and error calculation. Among them, data preprocessing plays an important role on the process of wind power forecasting since the inputs of the forecasting model would be sensitive to the quality of data. As a result, this paper presents a survey on the methods for wind-data processing. These methods aim to preprocess and extract suitable features from numerical weather prediction (NWP) wind speeds and measured wind power data. Finally, this paper used a case study to demonstrate the important of the preprocessing step on wind power forecasting.