A multitree approach for global solution of ACOPF problems using piecewise outer approximations

Journal Article · 09 June 2018 · Computers and Chemical Engineering

DOI: https://doi.org/10.1016/j.compchemeng.2017.10.018 · OSTI ID:1570257

Liu, Jianfeng ^[1]; Bynum, Michael ^[1]; Castillo, Anya ^[2]; Watson, Jean-Paul ^[2]; Laird, Carl D. ^[3]

---

1. Purdue Univ., West Lafayette, IN (United States). Davidson School of Chemical Engineering
2. Sandia National Lab. (SNL-NM), Albuquerque, NM (United States). Center for Computing Research
3. Purdue Univ., West Lafayette, IN (United States). Davidson School of Chemical Engineering; Sandia National Lab. (SNL-NM), Albuquerque, NM (United States). Center for Computing Research

Electricity markets rely on the rapid solution of the optimal power flow (OPF) problem to determine generator power levels and set nodal prices. Traditionally, the OPF problem has been formulated using linearized, approximate models, ignoring nonlinear alternating current (AC) physics. These approaches do not guarantee global optimality or even feasibility in the real ACOPF problem. We introduce an outer-approximation approach to solve the ACOPF problem to global optimality based on alternating solution of upper- and lower-bounding problems. The lower-bounding problem is a piecewise relaxation based on strong second-order cone relaxations of the ACOPF, and these piecewise relaxations are selectively refined at each major iteration through increased variable domain partitioning. Our approach is able to efficiently solve all but one of the test cases considered to an optimality gap below 0.1%. This approach opens the door for global solution of MINLP problems with AC power flow equations.

• Research Organization: Sandia National Laboratories (SNL-NM), Albuquerque, NM (United States)
• Sponsoring Organization: USDOE National Nuclear Security Administration (NNSA); USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR) (SC-21)
• Grant/Contract Number: AC04-94AL85000
• OSTI ID: 1570257
• Report Number(s): SAND2019--10860J; 679386
• Journal Information: Computers and Chemical Engineering, Journal Name: Computers and Chemical Engineering Journal Issue: C Vol. 114; ISSN 0098-1354
• Publisher: ElsevierCopyright Statement
• Country of Publication: United States
• Language: English