Performance Monitoring and Trend Visualization of Chillers Based on Principal Component Analysis and Its Application

Chillers are the key equipment in the heating, ventilation, and air conditioning (HVAC) system, which has a decisive impact on daily life and industrial production. However, since the operation and maintenance of chillers mainly rely on personnel experience, this may lead to problems such as the unreasonable setting of operating parameters, untimely troubleshooting of abnormalities, and degradation of unit performance, resulting in additional waste of energy. Therefore, researching performance monitoring and trend visualization of chillers significantly improves their operating efficiency. With big data technology’s rapid development and popularization, data-driven methods have become the research focus. However, most existing methods only focus on whether the chiller is faulty or abnormal and ignore the changes in the chiller’s performance, let alone visualize the changing trend. For this reason, this paper aims at the visualization of the performance trend of chillers and uses the principal component analysis (PCA) model to map the high-dimensional operating data to a two-dimensional space and then calculate the kernel density distribution of the data after dimensionality reduction, and constructs a base map that reflects chiller performance trends based on kernel density. In online monitoring, data representation in two-dimensional space is obtained by performing PCA dimension reduction processing on real-time data. Then, the polar coordinate transformation is displayed as points on the trend base map to realize the dynamic monitoring and trend visualization of chiller performance.