基于差异补偿和短期采样对比损失的城市电力负荷预测方法
Urban Electricity Load Forecasting Method Based on Discrepancy Compensation and Short-term Sampling Contrastive Loss
陈润桓 1戴华 2郑桂能 3李惠 1杨庚2
作者信息
- 1. 南京邮电大学计算机学院、软件学院、网络空间安全学院 南京 210023
- 2. 南京邮电大学计算机学院、软件学院、网络空间安全学院 南京 210023;江苏省大数据安全与智能处理重点实验室 南京 210023
- 3. 犹他州立大学计算机科学系 洛根84322
- 折叠
摘要
城市电力负荷预测是城市智能电网规划和调度的一项重要内容.然而,城市电力负荷预测中存在数据不均的问题,给城市电力负荷预测带来了巨大挑战.传统的基于单一模型的方法难以解决数据不均的问题,而现有的基于多模型的预测方法根据电力负荷分布将数据集拆分成多个子数据集,然后分别建立多个预测模型进行预测,该类方案在一定程度上解决了数据不均问题,但存在模型构建成本较高、不同分布样本间共有的电力分布特征发生分离等问题.基于此,提出了一个轻量级城市电力负荷预测模型(Lighten-DCSC-LSTM).该模型通过在长短期记忆网络的基础上引入差异补偿的思想和短期采样对比损失进行构建,同时构建共享特征提取层来降低模型构建成本.其中,差异补偿思想通过学习不同电力负荷分布样本之间的差异对主序列预测模块的预测结果进行差异补偿,短期采样对比损失通过动态类中心的对比学习损失对模型的训练进行正则化.为了验证模型的性能,进行了参数调优和对比实验.对比实验结果表明,模型在预测电力负荷的任务中取得了良好的性能.
Abstract
Urban power load forecasting is an important content of urban smart grid planning and scheduling.However,the pro-blem of data imbalance in urban power load forecasting poses a great challenge to urban power load forecasting.Traditional sin-gle-model-based methods can hardly solve the problem of data imbalance.The existing multi-model-based forecasting methods split the datasets into multiple sub-datasets according to the electricity load profiles,and then build multiple forecasting models for forecasting,which can solve the data imbalance problem to a certain extent,but there are problems such as high model con-struction cost and separation of the common electricity distribution characteristics among different distribution profiles.Based on this,this paper proposes a lighten urban electric load forecasting model(Lighten-DCSC-LSTM).It is constructed by introducing the idea of discrepancy compensation and short-term sampling contrastive loss on the basis of long and short-term memory net-works,while building a shared feature extraction layer to reduce the model construction cost.Among them,the idea of discrepan-cy compensation compensates the prediction results of the main sequence prediction module by learning the differences between different power load distribution samples,and the short-term sampling contrastive loss regularizes the training of the model by the contrastive learning loss of the dynamic class center.To verify the performance of the proposed model,parameter tuning and comparison experiments are conducted.The results of the comparison experiments show that the model achieves good perfor-mance in the task of forecasting electricity loads.
关键词
电力负荷预测/长短期记忆网络/深度学习/对比学习损失Key words
Electricity load forecasting/Long-short term memory networks/Deep learning/Contrastive learning loss引用本文复制引用
基金项目
国家自然科学基金(61872197)
国家自然科学基金(61902199)
国家自然科学基金(61972209)
安徽省高等学校科研计划重大项目(2022AH040148)
江苏省研究生科研创新计划(SJCX22_0265)
出版年
2024
NETL
NSTL
万方数据