摘要
为解决新能源发电功率间歇性和随机性问题,基于大气混沌特性展开对新能源电力系统与运行备用量化概率的预测研究,运用极限学习机分位数回归预测减小预测误差,保障系统运行安全。提出概率预测-决策一体化方案,围绕人工智能驱动,采用极限学习机分位数回归引入气象特征挖掘非参数区间预测方式对现有成本区间预测、约束条件等进行优化,设计集概率预测与决策一体的机器学习模型。研究结果显示,该方法有效克服了以往预测结果误差大、决策性能不足问题,降低了模型的复杂性,提高了管理协同性,保障了电力系统的安全、高效运行。
In order to solve the problems of intermittence and randomness of new energy power generation,the study conducts the prediction research on the new energy power system and operational reserve quantization probability based on the chaotic characteristics of the atmosphere.The extreme learning machine quantile regression prediction is used to reduce the prediction error and ensure the operation safety of the system.At the same time,an integrated scheme of probabilistic prediction and decision making is proposed.Driven by artificial intelligence,the extreme learning machine quantile regression is adopted to introduce meteorological features mining non-parametric interval prediction,optimize the existing cost interval prediction and constraint conditions,and design a machine learning model integrating probabilistic prediction and decision making.The results show that this method can effectively overcome the problems of large error and insufficient decision-making performance,reduce the complexity of the model,improve the coordination of management,and ensure the safe and efficient operation of the power system.
作者
谢云锋
Xie Yunfeng(Hebei Weizhou Energy Comprehensive Development Co.,LTD.,Zhangjiakou 075700,China)
出处
《黑龙江科学》
2024年第22期20-23,27,共5页
Heilongjiang Science
关键词
电力系统
备用量化
概率预测-决策
Power system
Reserve quantification
Probabilistic prediction-decision making
