Nonintrusive Load Monitoring for Verification and Diagnostics

Nonintrusive load monitoring (NILM) is a method of detecting the current energy consumption of a building, using a single set of sensors on the main building supply. This approach is in contrast to intrusive monitoring where end-use devices are sensed. Building on techniques of previous works, it will be shown that NILM can be implemented on commercially available devices with capabilities similar to modern smart meters and can provide meaningful feedback to both the user and supplying utility. Limitations of inexpensive commercial devices, such as resolution and measurement sample rate, will be addressed. Using clustering and a Hidden Markov Model approach, data about the state of the devices in a building can be determined. This information can then be used to verify the effectiveness of smart-grid initiatives such as VAr control and demand-side management in addition to other energy-saving measures such as weatherproofing and installing energy-efficient appliances. In addition, this information can be used to select devices for distributed control by analyzing not only the type of device but also the real-world operating characteristics