Stochastic Modeling and Integration of Plug-In Hybrid Electric Vehicles in Reconfigurable Microgrids With Deep Learning-Based Forecasting
Morteza Dabbaghjamanesh,Abdollah Kavousi-Fard,Jie Zhang
DOI: https://doi.org/10.1109/tits.2020.2973532
IF: 8.5
2021-07-01
IEEE Transactions on Intelligent Transportation Systems
Abstract:This paper investigates the impact of uncoordinated, coordinated, and smart charging of plug-in hybrid electric vehicles (PHEVs) on the optimal operation of microgrids (MGs) incorporating the dynamic line rating (DLR) security constraint. The DLR constraint, particularly in the islanding mode, influences the ampacity of MG feeders, when distribution lines reach their maximum capacity. To overcome any line outage or contingency situation, smart PHEVs are utilized to help improve the grid security. However, using PHEVs can cause higher power losses and feeder overloading issues. To address these concerns, a reconfiguration technique is employed in this paper. A heuristic algorithm, known as the collective decision-based optimization algorithm, is utilized to overcome the non-convexity and nonlinearity of the problem. The unscented transform technique is employed to model DLR uncertainties caused by solar radiation, load demand, and weather temperature, as well as PHEVs' uncertainties caused by varying charging strategies, numbers of PHEVs being charged, charging start time, and charging duration. Moreover, a deep learning gated recurrent unit technique is designed to forecast renewable power output for mitigating the uncertainties in renewable energy components. A modified IEEE 33-bus test network is deployed to evaluate the efficiency and performance of the proposed model.
engineering, electrical & electronic,transportation science & technology, civil