AC Transmission Network Expansion Planning: A Semidefinite Programming Branch-and-Cut Approach

Bissan Ghaddar,Rabih Jabr

DOI: https://doi.org/10.48550/arXiv.1711.03471

2017-11-10

Abstract:Transmission network expansion planning is a mixed-integer optimization problem, whose solution is used to guide future investment in transmission equipment. An approach is presented to find the global solution of the transmission planning problem using an AC network model. The approach builds on the semidefinite relaxation of the AC optimal power flow problem (ACOPF); its computational engine is a new specialized branch-and-cut algorithm for transmission expansion planning to deal with the underlying mixed-integer ACOPF problem. Valid inequalities that are based on specific knowledge of the expansion problem are employed to improve the solution quality at any node of the search tree, and thus significantly reduce the overall computational effort of the branch-and-bound algorithm. Additionally, sparsity of the semidefinite relaxation is exploited to further reduce the computation time at each node of the branch-and-cut tree. Despite the vast number of publications on transmission expansion planning, the proposed approach is the first to provide expansion plans that are globally optimal using a solution approach for the mixed-integer ACOPF problem. The results on standard networks serve as important benchmarks to assess the solution quality from existing techniques and simplified models.

Optimization and Control