HVAC (Heating, Venting, and Air Conditioning) System Fault Diagnosis and Energy Optimization Using State-Space Methods and Modern Control Theory

摘要

The report addresses the fault detection, diagnosis, and control of Heating, Venting, and Air Conditioning (HVAC) systems. The problems associated with the effective operation of HVAC systems are identified and a hierarchical framework for the operation of the system is formulated. A modern multivariable control approach, Model Algorithmic Control (MAC), was found to be a suitable approach for integrating the different hierarchies in the operation of the system. A methodology for detecting and diagnosing faults in the system using a nonlinear mathematical model of the system plus extended Kalman filtering was developed. This technique was implemented in a computer program and successfully used to detect 'planted' faults in the air handling unit of the HVAC system.