2019 · 11 citations · 20 references
Natural SciencesIndustrial EngineeringMathematicsDesign OptimizationEnergy EconomicsOptimizationCombinatorial OptimizationConvex OptimizationEnergy System ModelingInteger ProgrammingNetwork Flow AlgorithmsOptimization TheoryConvergenceConvex FunctionsGas FlowGas Network IntegrationMixed-integer ProgrammingNatural GasPenalized OptimizationSecond-order Cone Programming
Due to their environmental and economic benefits, interdependencies between power and gas systems have been enhanced. However, they provide new challenges in the optimal power-gas flow (OPGF) problem, especially at distribution-level. This paper proposes a sequential mixed-integer second-order cone programming (S-MISOCP) algorithm to define the OPGF for integrated electricity and gas system. The nonconvex power flow and gas flow equations are decomposed as difference-of-convex functions which are reformulated as MISOCP constraints. Starting with an initial point, a sequence of penalized MISOCP problems are solved to find a feasible OPGF close to, if not equal to, the optimal one. The feasibility and quick convergence are guaranteed by designing an adaptive penalty growth rate and suggesting a high-quality initial point, respectively. Moreover, bidirectional gas flow inside pipelines is considered to incorporate meshed gas networks. Numerical simulations have been conducted to validate the algorithm performance.
20