An Association Rule-Based Online Data Analysis Method for Improving Building Energy Efficiency
- 208 Downloads
Association rule mining has been applied to reveal variable relations from numerous operational data in buildings. However, there is a lack of effective methods to take full advantage of the discovered relations. This study proposes a real-time data analysis method for diagnosing building operational problems based on the discovered relations. In this method, the historical data are explored by the association rule mining to generate raw association rules. The abnormal and normal rules are extracted manually to build a rule base. The rule base is then used to analyze the real-time measurements of the relevant variables in the extracted rules. Operational problems are detected if the measurements of the relevant variables match with an abnormal rule or break all the related normal rules. Evaluations are made using the operational data collected from the chiller plant of a commercial building located in Shenzhen, China. Results show that the proposed method can detect operational problems effectively.
KeywordsData mining Association rule mining Expert system HVAC system Operational problem diagnosis
This study is supported by the National Nature Science Foundation of China (Number 51706197).