摘要
为了研究风功率预测的准确性对接纳区间确定的影响,进而提高电网调度的决策能力,结合长短期记忆(long short-term memory,LSTM)网络和注意力机制,提出了一种基于Attention-LSTM的风功率预测模型,并基于该预测模型进一步优化了电网风电接纳能力的评估方法。首先,利用风功率影响因素构建基于Attention-LSTM的风功率预测模型;其次,建立风电接纳能力评估模型;最后,使用Matlab进行仿真实验,对Attention-LSTM模型与传统模型的预测数据进行对比分析,并基于预测数据对电网风电接纳能力进行确定及评估。结果表明:Attention-LSTM模型预测数据的平均绝对误差为4.602 MW,在准确性上明显优于传统预测模型;风功率预测值与风电接纳区间预测值的相关性较高,提升风功率预测的准确性可以提高风电接纳区间上限的精确性。基于Attention-LSTM的风功率预测模型相较于传统预测模型精确度更高,在实际生产中能够保证电力系统稳定运行。
In order to study the influence of the accuracy of wind power prediction on the determination of the acceptance interval and improve the decision-making ability of grid dispatching,a wind power prediction model based on Attention-LSTM was proposed by combining the long short-term memory(LSTM)network and attention mechanism,and the evaluation method of wind power acceptance capacity of the grid was further optimized based on the prediction model.Firstly,the influencing factors of wind power were used to construct a wind power prediction model based on Attention-LSTM.Secondly,an evaluation model of wind power acceptance capacity was established.Finally,Matlab was used to carry out simulation experiments to compare and analyze the prediction data of the Attention-LSTM model and the traditional model,and to determine and evaluate the wind power acceptance capacity of the grid based on the prediction data.The results show that the average absolute error of the prediction data of the Attention-LSTM model is 4.602 MW,which is significantly better than that of the traditional prediction model,and the wind power prediction value has a high correlation with the wind power acceptance interval prediction,and the accuracy of the upper limit of the wind power acceptance interval can be improved by improving the accuracy of the wind power prediction accuracy.Compared with the traditional prediction model,the wind power prediction model based on Attention-LSTM has higher accuracy and can ensure stable operation of the power system in actual production.
作者
韩秀峰
佟金锴
陈卫东
车笛
赵翔宇
王亮
HAN Xiufeng;TONG Jinkai;CHEN Weidong;CHE Di;ZHAO Xiangyu;WANG Liang(State Grid Jilin Electric Power Company Limited Ultra High Voltage Company,Changchun,Jilin 130041,China;School of Electric Engineering,Shenyang Institute of Technology,Shenyang,Liaoning 110136,China)
出处
《河北工业科技》
2025年第3期285-294,共10页
Hebei Journal of Industrial Science and Technology
基金
辽宁省科技厅创新能力提升联合基金(2022-NLTS-16-05)
辽宁省教育厅基本科研项目(LJKMZ20221707)。
关键词
发电工程
风功率预测
风电接纳能力
注意力机制
弃风率
power generation engineering
wind power prediction
wind power acceptance capacity
attention mechanism
wind curtailment rate
