Stochasticity agnostic solution to the AC optimal power flow by recursive bound tightening with top‐down heuristically inducted binary decision trees

The efficient solution of the AC optimal power flow (OPF) for all cases and electrical grids is not guaranteed. Heuristics, approximations relaxations have been proposed aplenty, each with pros cons. This work proposes to solve OPF binary decision trees (BDTs). starts full feasible set an instance, and, iteratively, BDTs tighten constraints/bounds that set, improve occurring by median objective function preceding set. medians progressively tightened sets will converge global optimum OPF. Recent proofs estimate performance top-down BDT learning heuristics ensure convergence method optimum, provided adequately sampled training. recursive implementation inductive nature may also yield dispatches at multiple optimality levels same account stochastic resources. assessed over NICTA IEEE PES task force on benchmarks validation emerging system algorithms.