Development and Validation of Home Comfort System for Total Performance Deficiency/ Fault Detection and Optimal Comfort Control

This report explores the potential of using smart thermostat data for optimal control and performance degradation detection for home ACs. We developed and tested a learning-based home thermal model that facilitates the operation of a model predictive control (MPC)-based optimization agent and an automated fault detection and diagnosis (AFDD) agent. The home thermal model, along with the MPC-based optimization agent and AFDD agent, were tested in several test homes, as well as field tested in nine demonstration homes with real occupants. Ultimately, the home thermal model and its two parameter identification methods were successfully verified. For the MPC agent, field tests at the demonstration homes showed that with MPC, up to 51% and 62% cost savings can be obtained on hot and mild summer days, respectively. For the AFDD agent, laboratory and field tests showed it to be capable of catching less severe (around 15%) AFR faults, and the agent was overall found to be effective.