通信与信息技术Issue(2)：70-74,5.
基于张量分解的室内温度数据补全技术研究
Research on indoor temperature data completion technology based on tensor decomposition
摘要
Abstract
Indoor temperature is the fundamental principle for construction,operation,intelligent control,smart heating,energy con-servation and consumption reduction,daily management,technological progress,and satisfactory service.However,difficulties in obtain-ing room temperature data and equipment damage have resulted in missing room temperature,which poses challenges for real-time moni-toring and precise control,and leads to difficulties in accurately assessing and analyzing indoor thermal environments.In order to solve the problem of missing room temperature data,a tensor completion method based on Tucker Regularization is proposed using the concept of data completion.The missing room temperature dataset is completed by adding L2 norm to the core tensor to improve robustness,and the alternating least squares method is used to solve the model.In the study,real room temperature data was used to investigate how to im-prove completion accuracy in different missing scenarios.The research results show that the proposed algorithm outperforms the com-pared algorithms in terms of performance,and has good applicability and scalability under high missing rates.关键词
室内温度/缺失数据补全/张量补全/L2范数Key words
Indoor temperature/Missing data completion/Tensor completion/L2-norm分类
信息技术与安全科学引用本文复制引用
孙龙,李海广,秦杨,张超..基于张量分解的室内温度数据补全技术研究[J].通信与信息技术,2025,(2):70-74,5.基金项目
国家自然科学基金资助项目(项目编号:50965052) （项目编号:50965052）
