Disturbances can be added to an object to model breakdowns/failures. It is possible to add one or several disturbances, e.g. one for short machine failures, one for long machine failures and one for tool failures etc (see example in the top-right image). This can be useful if there are different failure behaviors with different origins that can't be combined in a common failure profile but on their own they have failure profiles that realistically represents the reality. For each disturbance the provided information decides the breakdown frequency and length, details about how to define a disturbance are given further down.
Note: Individual disturbances are independent, but as soon as (at least) one disturbance is failed the object will be failed. With other words all disturbances need to be operational (non-failed) for the object to be operational.
Editing several disturbances is done using the 'Edit Disturbances'-dialog
(top-right) this dialog is accessed using the
-button of the Disturbances setting
(green cicrle, in the image to the right). Each row represents
a disturbance and the following actions are used to add and remove
disturbances:
Add a disturbance: Push 
to add a disturbance.
Remove a disturbance: Push 
for the disturbance that should be removed.
Remove all disturbances (except the first one): Push
in the header row. There must be a disturbance defined, however it can be
one with 100 % availability.
Note: The Multi-select feature will delete (not without confirmation) any opitmization inputs that are linked to the disturbances that are being edited. However, editing disturbances of individual objects will not.
Defining a disturbance
A disturbance can be defined in one out of three ways:
- Percent: Define a disturbance by specifying an availablility and a mean time to repair (mttr).
- Distributions: Define a disturbance by specifying separate distributions for the duration of failures (duration) and the time between failures (interval).
-
Cycles: Define a disturbance by specifying separate
distributions for the duration of failures (duration) and the number of
entities that are processed between failures (interval).
Can be used to introduce losses that occur with a certain frequency, e.g. a tool change after every 1000th processed entity.
Note: The cycle count (i.e. the number of processed entities) are randomly initialized between 1 and the mean value of the statistical distribution used for the interval.
The input-field of the disturbances setting contain a string representing all sub-settings of all defined disturbances. The sub-settings of the first added disturbance are shown by expanding the disturance setting using the >-symbol to the left of the disturbances setting. The list below gives more details on the sub-settings of the disturbances setting.
Note: Selecting a certain value of a sub-setting can disable or restrict the options of one or several other sub-settings.
|
Sub-setting |
Description |
|
Name |
A name for the disturbance. This must be unique (i.e. no two or more disturbances on the same object is allowed to have the same name) since this is used to distinguish between the different disturbances. |
|
Type |
Percent: (DEFAULT) The disturbance will be spcified with an Availability and a Mttr.
Distributions: The disturbance will be specified by the two distributions Duration and Interval.
Cycles: The disturbance will be specified by the two distributions Duration and Interval.
|
|
Availability |
A percentage that specifies the amount of the time that the object will be operational (not failed). Range of values ]0,100], default 100 %.
|
|
Time reference |
Can be set to either of {simulation,operational,processing} and it will determine how the time between two failures will be counted. If simulation is selected If operational is selected If processing is selected Simulation: All time no matter the state of the object will be counted as time between failures. Operational: The time between two failures is only counted when the object is operational (i.e. not failed or paused). Processing: The time between two failures is only counted when the object is processing something. (For Stores this will likely result in lower availability than expected.) |
|
Duration |
A distribution that randomly sets the time it takes to get an object operational again after a failure.
|
|
Interval |
A distribution that randomly sets the time to the next failure once the previous failure have ended.
|
|
Mttr (Mean Time To Repair) |
The average time it will take to get the object operational again when it breaks down. Default value is 0 seconds.
|