Enhancing large neighbourhood search heuristics for Benders’ decomposition

Abstract A general enhancement of the Benders’ decomposition (BD) algorithm can be achieved through improved use large neighbourhood search heuristics within mixed-integer programming solvers. While solvers are endowed with an array heuristics, few, if any, have been designed for BD. Further, typically is limited to finding solutions BD master problem. Given lack frameworks BD, only ad hoc approaches developed enhance ability find high quality primal feasible heuristics. The framework SCIP has extended a trust region based heuristic and employs solve auxiliary problems all improve identified solutions. computational results demonstrate that technique accelerate improvement in bound when applying