区域风电场群集中式功率预测系统设计  

Design of a Centralized Power Prediction System for Regional Wind Farm Clusters

作　　者：何祥针 左剑 杨韵 付聪 刘锋 He Xiangzhen;Zuo Jian;Yang Yun;Fu Cong;Liu Feng(Power Dispatch and Control Center,Guangdong Power Grid Co.,Ltd.,Guangzhou Guangdong 510220,China;Beijing Tsintergy Technology Co.,Ltd.,Beijing 100084,China)

机构地区：[1]广东电网有限责任公司电力调度控制中心,广东广州510220 [2]北京清能互联科技有限公司,北京100084

出　　处：《电气自动化》2024年第5期59-61,65,共4页Electrical Automation

基　　金：国家自然科学基金项目(GDKJXM20210069)。

摘　　要：为实现对区域风电场群集中式功率的准确预测,提出并设计了一种适应性强、维度高的预测模型。首先,通过AHP分析法发现影响区域风电场群集中式功率预测系统的五个必要因子;然后,构建预测评估模型得到其设计附加值的数据构建状态空间模型;接着,将该状态空间模型在人工智能的数据空间中进行部署并加以神经网络训练;最后,对这一预测系统模型进行实证验证,发现其数据吻合度较高。该预测模型可以用于对区域风电场群集中式功率进行预测,具有较广泛的应用价值。In order to accurately predict the centralized power of regional wind farm clusters,a prediction model with strong adaptability and high dimensionality was proposed and designed.Firstly,five essential factors affecting the centralized power prediction system of regional wind farm clusters were identified through AHP analysis;then,a predictive evaluation model was constructed to obtain the data of its design added value,and a state space model was built;next,the model was deployed in the data space of artificial intelligence and trained with neural networks;finally,the empirical verification was conducted on the prediction system model.It has been found that the model has high data consistency and can be used to predict the concentrated power of regional wind farm clusters,which has broad application value.

关 键 词：区域风电场群 集中式功率 预测评估模型 神经网络训练 状态空间模型 

分 类 号：TM425[电气工程—电器]