| 引用本文: | 崔树银,汪昕杰.基于最大信息系数和多目标Stacking集成学习的综合能源系统多元负荷预测[J].电力自动化设备,2022,42(5): |
| CUI Shuyin,WANG Xinjie.Multivariate load forecasting in integrated energy system based on maximal information coefficient and multi-objective Stacking ensemble learning[J].Electric Power Automation Equipment,2022,42(5): |
|
| 摘要: |
| 精确的多元负荷预测对于综合能源系统的能源调度与运行规划起到重要的作用。对电、热、冷负荷单独进行预测的传统方法会忽略多元负荷间的耦合关系。针对这一问题，提出一种基于多目标Stacking集成学习的多元负荷协同预测模型。引入最大信息系数对多元负荷及天气因素进行相关性分析，并提出负荷耦合形态指标来深度挖掘多元负荷间的耦合关系；将多目标回归与Stacking集成学习模型相结合，建立多元负荷协同预测模型；通过实际算例验证所提模型的有效性，算例结果表明，与其他预测模型相比，所提模型预测精度更高。 |
| 关键词: 多目标回归 Stacking集成学习 综合能源系统 最大信息系数 正则化贪心森林算法 |
| DOI：10.16081/j.epae.202202025 |
| 分类号:TM715 |
| 基金项目:国家自然科学基金资助项目(71972127) |
|
| Multivariate load forecasting in integrated energy system based on maximal information coefficient and multi-objective Stacking ensemble learning |
|
CUI Shuyin, WANG Xinjie
|
|
School of Economics and Management, Shanghai University of Electric Power, Shanghai 200090, China
|
| Abstract: |
| The accurate multivariate load forecasting plays an important role in energy dispatching and operation planning of integrated energy system. Traditional methods for forecasting electrical, thermal, and cooling loads separately ignore the coupling relationship between multivariate loads. In order to solve this problem, a multivariate load collaborative forecasting model based on multi-objective Stacking ensemble learning is proposed. The maximal information coefficient is introduced to analyze the correlation between multivariate load and weather factors, and the load coupling morphological index is proposed to deeply explore the coup-ling relationship between multivariate loads. Then, the multi-objective regression and Stacking ensemble learning model are combined to establish the multivariate load cooperative forecasting model. A practical example is given to verify the effectiveness of the proposed model, and the results of numerical examples show that the forecasting accuracy of the proposed model is higher than that of other forecasting models. |
| Key words: multi-objective regression Stacking ensemble learning integrated energy system maximal information coefficient regularized greedy forest algorithm |
