[R] Help to simulate data to model pump failure use case

Sunny Singha sunnysingha.analytics at gmail.com
Fri Jun 16 07:18:50 CEST 2017


Hi,
I need help to simulate data to model pump's failure. Below is thought
process to simulate the data:
- For each Pump readings are captured via sensors each minute for period of
3 months. There are 10 pumps in total.
- The failure rate or occurrences of event in 3 months for each Pump is
defined. i.e Pump1 would fail 3 times in 3 months, Pump2 would fail 2 times
in 3 months, etc.
- The failure modes are defined based on the combination of below variables:
   Failure modes::
   -- Bearing damage --> +ve_corr('temprature' + vibration)
   -- Cavitation --> -ve_corr( +ve_corr(discharge_pressure + flow) +
temprature)
   -- worn shafts --> +ve_corr(discharge_pressure + flow)

How should I be generate distributions for the variables listed in the
"failure modes" maintaining the correlation stated above.

Regards,
Sunny

	[[alternative HTML version deleted]]



More information about the R-help mailing list