Job Shop Scheduling with Integer Programming, Shifting Bottleneck, and Decision Diagrams: A Computational Study

TL;DR

This paper compares classical and novel heuristic algorithms for job shop scheduling, including decision diagrams and local refinement, through computational experiments to evaluate their effectiveness.

Contribution

It introduces novel strategies using decision diagrams and combines heuristics with MIP and CP approaches for improved job shop scheduling solutions.

Findings

Decision diagrams enhance heuristic performance.

Balas' local refinement improves solution feasibility.

Computational experiments demonstrate the effectiveness of combined methods.

Abstract

We study heuristic algorithms for job shop scheduling problems. We compare classical approaches, such as the shifting bottleneck heuristic with novel strategies using decision diagrams. Balas' local refinement is used to improve feasible solutions. Heuristic approaches are combined with Mixed Integer Programming and Constraint Programming approaches. We discuss our results via computational experiments.