This content is not included in your SAE MOBILUS subscription, or you are not logged in.
Artificial Intelligence-Based Performance Optimization of Electric Vehicle-to-Home (V2H) Energy Management System
ISSN: 2640-642X, e-ISSN: 2640-6438
Published September 22, 2020 by SAE International in United States
Citation: Shamami, M., Alam, M., Ahmad, F., Shariff, S. et al., "Artificial Intelligence-Based Performance Optimization of Electric Vehicle-to-Home (V2H) Energy Management System," SAE J. STEEP 1(2):115-125, 2020, https://doi.org/10.4271/13-01-02-0007.
Electric vehicles (EVs) with enormous batteries are capable of storing a high quantity of energy. This stored energy can be used as a standby power supply or as backup energy storage for domestic loads whenever a blackout or load shedding happens. Further, solar photovoltaic (PV) energy can also be utilized to charge the battery of EV and also improve the backup energy source for domestic loads. The main purposes of this article are to reduce the energy cost of a household and also to minimize the dependency of domestic loads on the grid, to enhance the reliability of power supply to the residential loads during load shedding and blackouts, and to maximize the utilization of power produced by the solar photovoltaic array (SPVA) mounted on the rooftop.
Moreover, Alternating Current (AC) loads are connected to the AC bus, and Direct Current (DC) loads are connected separately to the DC bus to avoid power losses in conversion (DC to AC and AC to DC). In this work, a fuzzy inference system (FIS) is applied for effective energy management of a home and better utilization of stored energy of EVs for domestic loads. A model of vehicle-to-home (V2H) system based on the fuzzy logic controller (FLC) for managing the battery of the vehicle, rooftop SPVA, emergency backup power, DC, and AC residential loads, and grid power is also proposed. An effective and reliable FLC based on the Sugeno inference system has been designed and developed, which is optimized by an adaptive neuro-fuzzy inference system (ANFIS) for better performance. The number of rules for the first design was 47, which is reduced to 38 by ANFIS optimization. Therefore, 54 rules for six FLC have been reduced for the entire system. Further, the results show appreciable values in terms of economic parameters.