TY - JOUR
T1 - ADMM-Based Multiperiod Optimal Power Flow Considering Plug-In Electric Vehicles Charging
AU - Fan, Hua
AU - Duan, Chao
AU - Zhang, Chuan Ke
AU - Jiang, Lin
AU - Mao, Chengxiong
AU - Wang, Dan
N1 - Publisher Copyright:
© 1969-2012 IEEE.
PY - 2018/7
Y1 - 2018/7
N2 - When plug-in electric vehicles (PEVs) participate in grid operation, the intertemporal feature of PEVs charging transforms the traditional optimal power flow (OPF) problem into multiperiod OPF (MOPF) problem. In the case that the population of PEVs is huge, the large number of variables and constraints renders the centralized solution technique unsuitable to solve the MOPF problem. Therefore, a distributed algorithm based on alternating direction method of multipliers is developed to decompose the MOPF into two update steps that are solved in an alternating and iterative style. To improve the solution efficiency, the second update step is transformed into a Euclidean projection problem by approximating the original objective with a surrogate function. Then, a projection algorithm is utilized to solve the approximate problem. Numerical results show that this reformulated model obtains suboptimal solutions with small relative error, but gains considerable speed-up. Furthermore, its scalability and effectiveness are tested in the 119-bus and 906-bus distribution networks.
AB - When plug-in electric vehicles (PEVs) participate in grid operation, the intertemporal feature of PEVs charging transforms the traditional optimal power flow (OPF) problem into multiperiod OPF (MOPF) problem. In the case that the population of PEVs is huge, the large number of variables and constraints renders the centralized solution technique unsuitable to solve the MOPF problem. Therefore, a distributed algorithm based on alternating direction method of multipliers is developed to decompose the MOPF into two update steps that are solved in an alternating and iterative style. To improve the solution efficiency, the second update step is transformed into a Euclidean projection problem by approximating the original objective with a surrogate function. Then, a projection algorithm is utilized to solve the approximate problem. Numerical results show that this reformulated model obtains suboptimal solutions with small relative error, but gains considerable speed-up. Furthermore, its scalability and effectiveness are tested in the 119-bus and 906-bus distribution networks.
KW - Plug-in electric vehicles
KW - alternating direction method of multipliers (ADMM)
KW - multiperiod optimal power flow
KW - projection algorithm
UR - https://www.scopus.com/pages/publications/85039796745
U2 - 10.1109/TPWRS.2017.2784564
DO - 10.1109/TPWRS.2017.2784564
M3 - 文章
AN - SCOPUS:85039796745
SN - 0885-8950
VL - 33
SP - 3886
EP - 3897
JO - IEEE Transactions on Power Systems
JF - IEEE Transactions on Power Systems
IS - 4
ER -