Skip to content

Releases: hoff-diesel/simantha

v0.0.3 - Bug fixes, notation changes, other updates

06 Aug 18:13
Compare
Choose a tag to compare

This release fixes several bugs that occurred due to parallel SimPy processes. Occasionally, a repaired machine would wait one time unit before resuming production.

Machines are now indexed by i and the number of machines is n, to be consistent with related literature.

Removed graphviz dependency for now.

Some basic tests are included in tests.py.

v0.0.2 - Packaging and degradation improvements

05 Jun 20:04
Compare
Choose a tag to compare

The project has been packaged at https://pypi.org/project/maintsim and can now be installed using pip install maintsim.

This release also introduced improvements to degradation, allowing for more flexible degradation profiles. A system can be defined with complete degradation transition matrices for each machine. See the updated documentation for details and options.

Further testing to verify system behavior and validate simulation data collection is underway. Pending releases will also include improvements to efficiency for iterated simulations.

v0.0.1 - Initial release

15 May 17:49
Compare
Choose a tag to compare
Pre-release

This initial version has support for corrective and condition-based maintenance policies for a serial production line of machines that are subject to discrete-state Markovian degradation.

Introduced in this release are tag-based scheduling and the the Scheduler class. This approach no longer uses a SimPy Resource object to model maintenance resources and instead tags machines for maintenance requests and assignment as appropriate. The advantage of this new method is that it is easier to implement more advanced methods of scheduling since we no longer need to worry about updating the simpy.Resource queue.

The Scheduler class can use any method of resolving maintenance conflicts. See the updated documentation for details.

Upcoming improvements include validation of simulation data collection and more efficient simulation replication.