Automatic Continuous Commissioning of Measurement Instruments in Air Handling Units

Abstract

This paper presents a robust strategy based on a condition-based adaptive statistical method for automatic commissioning of measurement instruments typically employed in air-handling units (AHU). The multivariate statistic method, principal component analysis (PCA), is adopted and modified to monitor the air handling process. Two PCA models are built corresponding to the heat balance and pressure-flow balance of the air-handling process. Sensor faults can be detected and isolated using the Q-statistic and the Q-contribution plot. The fault isolation ability against typical component faults is improved using knowledge-based analysis. A novel condition-based adaptive scheme is developed to update the PCA models with the operation conditions for continuous online application. A commissioning tool is developed to implement the strategy. Simulation tests and field tests in a building in Hong Kong were conducted to validate the automatic commissioning strategy for typical AHU. The integration of the tool with a building management system (BMS) and its application is demonstrated.