D-25 計測データと経験的知識による建築環境システムの異常検知と診断の方法 : クラスター分析による異常検知とアスペクト推論法による診断

概要

論文の詳細を見る
The objective of this study is to propose a method of detecting and diagnosing faults in an architectural environmental system such as the combination of a building,the HVAC system,and the lighting system. Our final goal is to establish a method of constructing new rules for detection and diagnosis, which is not defined in advance, on-the-fly from empirical data and knowledge. This paper describes the results of preparatory studies towards a robust and reliable method. Cluster analysis is employed for fault detection and the aspect inference method is employed for fault diagnosis. We confirm that the combination the proposed method works well with artificial test data for fault detection and diagnosis.
社団法人空気調和・衛生工学会の論文
2007-08-24