Pyscheduling

Thu, 29 Sep 2022 10:39:58 +0000

Pyscheduling is an open-source python package to solve scheduling problems. The categories tackled are : Single Machine, Parallel Machines, Flowshop and Jobshop.

The infrastructure for each category is implemented and open to extension to accept problems with more specificity in terms of constraints and multi-objective opptimization. There are methods going from exact methods, heuristics, metaheuristics, B&B, …etc to solve the different problems of the given category.

Single Machine Scheduling Problems :

Minimize Total Weighted Lateness.
Minimize Total Weighted Lateness with release dates.
Minimize Total Weighted Lateness with sequence dependent setup time.
Minimize Total Weighted Lateness with release dates and sequence dependent setup time.
Minimize Maximal Lateness with release dates and precedence constraints.
Minimize Total Weighted Completion Time.
Minimize Total Weighted Completion Time with release dates.
Minimize Maximal Completion Time with sequence dependent setup time.
Minimize Maximal Completion Time with release dates and sequence dependent setup time.

Parallel Machine Scheduling Problems :

Minimize Maximal Completion Time with sequence dependent setup time.
Minimize Maximal Completion Time with release dates and sequence dependent setup time.

Flowshop :

Minimize Maximal Completion Time.
Minimize Maximal Completion Time with sequence dependent setup time.

Jobshop :

Minimize Maximal Completion Time.

An easy-to-use interface is available for both single and parallel machines problems sheduling. In addition to both interfaces, a benchmark module is also available to allow it for users and especially researchers to test and benchmark their implemented methods based on our infrastructure with a given instances benchmark set.

The package is open to contribute for anyone in order to enrich the categories infrastructure, to implement state-of-the-art methods and to tackle more constrained problems.

This work couldn’t have been without the participation of the awesome colleagues:

Younes Mimene
Karima Benatchba
Farouk Yalaoui

Efficient heuristics and metaheuristics for the unrelated parallel machine scheduling problem with release dates and setup times

Wed, 13 Jul 2022 09:00:02 +0000

In order to study the unrelated parallel machine scheduling problem with release dates and setup times we generated a new benchmark of 1620 instances, divided into three sets (Small, Medium and Large instances).

This benchmark was first introduced in the paper “Efficient heuristics and metaheuristics for the unrelated parallel machine scheduling problem with release dates and setup times” published in GECCO 2022, Boston. You can find the generation protocol details in the paper here.

The following Table shows the summary of the benchmark.

You can download the instances from here. You will need the free software compressor 7zip to open the file.

Efficient heuristics and metaheuristics for the unrelated parallel machine scheduling problem with release dates and setup times

Sat, 14 May 2022 23:20:43 +0000

Metaheuristics | Scheduling-cc

Pyscheduling

Efficient heuristics and metaheuristics for the unrelated parallel machine scheduling problem with release dates and setup times

Efficient heuristics and metaheuristics for the unrelated parallel machine scheduling problem with release dates and setup times