Which benchmarks exist to evaluate the performance of algorithms for Job-Shop or Flow-shop scheduling problems?
Mainly, there exist 3 benchmarks to test shop scheduling problems.
Namely they are Taillard, Structured and ORLib benchmarks. These benchmarks have different goals. The Taillard benchmark is the most used benchmark in the literature. The benchmark targets permutation flowshop, flowshop, open shop and job shop scheduling problems. For details and download of the benchmark refer to http://www.lifl.fr/~liefooga/benchmarks/benchmarks/
The second benchmark shop scheduling problems is the structured benchmark. As the name suggests, the goal of the benchmark is to have test instances having a structure additional to the test instances with random instances. It contains all the instances from the Taillard benchmark. Further details and the benchmark can be found at http://www.cs.colostate.edu/sched/generator/
The third benchmark focuses on collecting test cases found difficult to solve in industry. We refer to this benchmark as ORLib benchmark. Contains instances from Taillard + some instances from industry. More details and benchmarks can be found at http://people.brunel.ac.uk/~mastjjb/jeb/orlib/flowshopinfo.html