Open Collections

Optimization of energy consumption schedule of residential loads and electric vehicles

University of British Columbia

In the current electrical grid, utility companies have begun to use demand side management (DSM) programs and time-of-use (TOU) pricing schemes to shape the residential load profile. However, it is difficult for the residential users to respond to the pricing signal and manually manage the operation of various household appliances. Hence, the autonomous energy consumption scheduling of residential loads and electric vehicles (EVs) is necessary for the users to benefit from the DSM programs. In this thesis, we propose different algorithms to schedule the energy consumption of residential loads and EVs, and provide ancillary services to the electrical grid. First, we study the DSM for areas with high photovoltaic (PV) penetration. Since many rooftop PV units can be integrated in the distribution network, the voltage rise issue occurs when the reverse power flow from the households to the substation is significant. We use stochastic programming to formulate an energy consumption scheduling problem, which takes into account the voltage rise issue and the uncertainty of the power generation from PV units. We propose an algorithm by solving the formulated problem and jointly shave the peak load and reduce the reverse power flow. Subsequently, we study using the EVs to provide the frequency regulation service. We formulate a problem to schedule the hourly regulation capacity of the EVs using the probabilistic robust optimization framework. Our formulation takes into account the limited battery capacity of the EVs and the uncertainty of the automatic generation control (AGC) signal. An efficient algorithm is proposed to solve the formulated problem based on duality. Last but not least, we study the market participation of an aggregator which coordinates a fleet of EVs to provide frequency regulation service to an independent system operator (ISO). The two-settlement market system (i.e., the day-ahead market (DAM) and real-time market) is considered. We analyze two types of DAMs based on the market rules of New York ISO and California ISO. We formulate a problem to determine the bid for the aggregator in the DAM using stochastic program and conditional value at risk. Efficient algorithms are proposed to tackle the formulated problem.

Text

2016-06-21

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/58302

Electrical and Computer Engineering

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/