[关键词]
[摘要]
本文对比了最小二乘多元线性回归、岭回归、Lasso回归、随机森林、XGBoost在地下车站通风空调、垂直交通能耗预测领域的应用效果。研究发现,对于地下车站通风空调和垂直交通能耗预测,各算法的CV-RMSE均在10%以下,均可达到工程应用要求的精度。其中,XGBoost算法在通风空调能耗预测中的CV-RMSE为5.1%,在垂直交通能耗预测中的CV-RMSE为5.4%,预测效果明显优于其他算法。从计算成本来看,最小二乘回归、岭回归、Lasso回归算法计算成本较低,随机森林和XGBoost模型调参复杂、计算成本较高。本文对比了常用的数据驱动算法的在地下车站能耗预测中的预测精度和计算成本,为地下车站模型的搭建提供了算法参考。
[Key word]
[Abstract]
This paper has compared a variety of commonly-used data-driven methods in the field of energy prediction model of the ventilation and air-conditioning (VAC) as well as vertical transport (TRANS) in subway stations, including least square multiple linear regression, Ridge regression, Lasso regression, random forest, XGBoost. It is found that the CV-RMSE indices of the afore-mentioned methods are less than 10% for the VAC and TRANS energy prediction, the accuracy of which are adequate for the engineering application. Among them, the CV-RMSE of XGBoost is only 5.1% for the VAC model and 5.4% for the TRANS model, which is obviously better than other algorithms. From the perspective of computational cost, the least square regression, Ridge regression and Lasso regression algorithms have comparatively low cost, while the random forest and XGBoost models have high computational cost. In this paper, by comparing the prediction accuracy and calculation cost of the commonly-used data-driven methods, the algorithm reference for the development of subway station energy prediction model is provided.
[中图分类号]
U231.1
[基金项目]
“十三五”国家重点研发计划课题 公共交通枢纽建筑节能关键技术与示范项目(2018YFC0705006)