We present the ColoredPower algorithm, which is designed to provide collaborative electricity demand shaping for residential and small-business customers. Demand shaping for this market sector is an important and challenging problem, since the vast number of such customers collectively account for a large fraction of total electricity consumption, yet each individual's consumption is small. Under the PACEM system, customers participate by "coloring" their appliances with a qualitative priority such as "can be shut off at peak power." Demand shaping for this system must be scalable to millions of appliances, operate quickly and fairly across customers, and act on any given appliance infrequently. This last constraint is particularly challenging: if an appliance that switches on or off must not be switched again for many minutes, then at any instant, a large fraction of appliances may not be controllable. The ColoredPower algorithm addresses these challenges usin...
Vinayak V. Ranade, Jacob Beal