华中科技大学学报（自然科学版）2025，Vol.53Issue(2)：120-126,7.DOI:10.13245/j.hust.250794
同时发生设计日集合覆盖挑选与预测方法
Selection and prediction methods for coincident design day based on set covering problem
摘要
Abstract
In order to get coincident design day to accurately calculate the design cooling load of air-conditioning system,a method based on set covering problem was proposed in this study to select and predict coincident design day.In this method,the representation vectors between design days and room samples were generated through long-term hourly weather data and cooling load calculation models,and the selecting problem of coincident design days was transformed into solving set covering problem by set covering model.The ant colony algorithm was used to obtain an optimal set which contains several coincident design days.XGBoost algorithm was used to construct the mapping relationship between the room characteristic parameters and the elements of the optimal set.The coincident design day of a room in design was predicted from the optimal set.This method was applied to obtain the optimal set of coincident design days for Hong Kong and predict the design days of the rooms in design.The results in case study show that a coincident design day with accurate load calculation can be selected from the optimal set for any characteristic-parameters-combination rooms.The difference between the design cooling load calculated by the predicted coincident design day and the theoretical design cooling load is less than 1.5%,which meets the accuracy requirement in engineering application.关键词
同时发生设计日/集合覆盖问题/最优集合/XGBoost算法/设计冷负荷Key words
coincident design day/set covering problem/optimal set/XGBoost algorithm/design cooling load分类
建筑与水利引用本文复制引用
陈友明,谢广仁..同时发生设计日集合覆盖挑选与预测方法[J].华中科技大学学报（自然科学版）,2025,53(2):120-126,7.基金项目
国家自然科学基金资助项目(52130802). （52130802）
