Convex Polar Second-Order Taylor Approximation of AC Power Flows: A Unit Commitment Study

Modern mixed-integer quadratic solvers generally handle binary variables more efficiently than nonlinear solvers. This is relevant to the power system operation models as unit commitment formulations typically contain a large number of variables. paper investigates how achieve accuracy level close one exact models, but by utilising convex and The presented model based on Taylor-series expansion where both voltage magnitude angle are quadratically constrained. To high accuracy, takes advantage meshed transmission network structure that enables replacement inequality constraints cause constraint relaxation errors with linear equality constraints. Quadratic be replaced well operating point parameters determined presolve. first case study validates model's convergence iterative algorithm, while second non-iterative full problem. Unit results show superior similar computation times existing hand faster polar formulation other.