Welcome Guest! To enable all features please Login. New Registrations are disabled.

Notification

Icon
Error

Login


2 Pages<12
Options
Go to last post Go to first unread
Offline Razonar  
#21 Posted : 12 September 2018 07:38:15(UTC)
Razonar


Rank: Advanced Member

Groups: Registered
Joined: 28/08/2014(UTC)
Posts: 1,356
Uruguay

Was thanked: 815 time(s) in 516 post(s)
Hi. NM algorithm with restrictions.
Best regards.
Alvaro

Weibull3.sm (50kb) downloaded 20 time(s).
Offline - Norm -  
#22 Posted : 12 September 2018 22:54:36(UTC)
- Norm -


Rank: Advanced Member

Groups: Registered
Joined: 29/11/2017(UTC)
Posts: 82
United States
Location: Midwest

Was thanked: 9 time(s) in 7 post(s)
Thanks Alvaro!

Interesting. The data's outcome can be configured to look based on the limit: beta>1 or beta<1. Which one is correct? More for me to ponder...

Norm
Offline Razonar  
#23 Posted : 13 September 2018 02:31:09(UTC)
Razonar


Rank: Advanced Member

Groups: Registered
Joined: 28/08/2014(UTC)
Posts: 1,356
Uruguay

Was thanked: 815 time(s) in 516 post(s)
Originally Posted by: nasch Go to Quoted Post
Thanks Alvaro!

Interesting. The data's outcome can be configured to look based on the limit: beta>1 or beta<1. Which one is correct? More for me to ponder...

Norm


Hi Norm. You can see Weibull not as a probability distribution but a family of them. When you choose a probability distribution is because someone else says that it holds for your project because he study a lot of problems seeing which one is the"correct". Question about which model is better is about which model is preferable for your 230 units: the beta < 1 sub family or the other. This information usually is given by the units manufacturer. Hope this helps.

Best regards.
Alvaro.

Offline Jean Giraud  
#24 Posted : 14 September 2018 07:17:08(UTC)
Jean Giraud

Rank: Guest

Groups: Registered
Joined: 04/07/2015(UTC)
Posts: 6,868
Canada

Was thanked: 980 time(s) in 808 post(s)
Originally Posted by: nasch Go to Quoted Post
Interesting. The data's outcome can be configured to look based on the limit: beta


More I read this more confused I get:

The life cycle analysis starts with n=130 units under test for time_test=499 days.
Several units fail. What is the probability that the ith unit will fail at time T hours.
Use a Weibull probability destiny distribution function. The family of rate distribution
functions is given by T: Time to failure in days.


As it looks to me: it's an exercise from undocumented lecture, thus expressed in terms of
incomplete understanding of what's needed for expressing results.
Your explanation just says: failures were observe over 499 days observed @ T days.
The missing link seems which one of the 130 unit failed ?

Stat Treasury_8 Weibull FailureAnalysis] Copy.sm (23kb) downloaded 5 time(s).

Offline Jean Giraud  
#25 Posted : 14 September 2018 07:55:46(UTC)
Jean Giraud

Rank: Guest

Groups: Registered
Joined: 04/07/2015(UTC)
Posts: 6,868
Canada

Was thanked: 980 time(s) in 808 post(s)
Originally Posted by: Jean Giraud Go to Quoted Post
More I read this more confused I get:


... more to read to de-confuse myself.

When reading papers, if I don't see the step by step of the mathematical analysis,
reproducible steps by steps on CAS, the writer writes for the birds.

Stat Treasury_8 Weibull FailureAnalysis] Copy.sm (30kb) downloaded 8 time(s).

Offline Razonar  
#26 Posted : 14 September 2018 08:29:12(UTC)
Razonar


Rank: Advanced Member

Groups: Registered
Joined: 28/08/2014(UTC)
Posts: 1,356
Uruguay

Was thanked: 815 time(s) in 516 post(s)
Originally Posted by: Jean Giraud Go to Quoted Post
Originally Posted by: Jean Giraud Go to Quoted Post
More I read this more confused I get:



Hi Jean. Units fails, but not dead nor disappear. They can be replaced or repaired. Norma's calculus is for planning the maintenance of the units for ensure some minimal critical number of them. For sure you can't allow only 1.47 units after 200 days, you do something for prevent that, this calculus just shows when you must to do that something.

Depending of the kind of "units", this calculus also are for not overload the workshop, have enough replacements at the warehouse, ensure the reliability of the plant, and a long etc.

Best regards.
Alvaro.
Offline - Norm -  
#27 Posted : 14 September 2018 21:40:22(UTC)
- Norm -


Rank: Advanced Member

Groups: Registered
Joined: 29/11/2017(UTC)
Posts: 82
United States
Location: Midwest

Was thanked: 9 time(s) in 7 post(s)
Additional reading... A pdf that was linked from a MC file using a different method to determine beta and the SMATH translated worksheet.

Norm

Interestingly, using the method to determine beta on the "chain analysis" sheet figures a beta 0.77, much less than all the methods used earlier in this forum-topic-chain. kmath122.pdf (510kb) downloaded 13 time(s). Weibull Widgets.sm (72kb) downloaded 12 time(s).
Offline mkraska  
#28 Posted : 14 September 2018 21:55:05(UTC)
mkraska


Rank: Advanced Member

Groups: Registered
Joined: 15/04/2012(UTC)
Posts: 1,980
Germany

Was thanked: 1120 time(s) in 718 post(s)
Originally Posted by: nasch Go to Quoted Post
Additional reading... A pdf that was linked from a MC file using a different method to determine beta and the SMATH translated worksheet.

Norm

Interestingly, using the method to determine beta on the "chain analysis" sheet figures a beta 0.77, much less than all the methods used earlier in this forum-topic-chain. kmath122.pdf (510kb) downloaded 13 time(s). Weibull Widgets.sm (72kb) downloaded 12 time(s).


The Maxima-based fit in post #9 gave a value of 0,8 which is not that far off 0,77.

Which, of course, doesn't mean anything.
Martin Kraska

Pre-configured portable distribution of SMath Studio: https://smath.com/wiki/SMath_with_Plugins.ashx
Offline - Norm -  
#29 Posted : 15 September 2018 00:27:23(UTC)
- Norm -


Rank: Advanced Member

Groups: Registered
Joined: 29/11/2017(UTC)
Posts: 82
United States
Location: Midwest

Was thanked: 9 time(s) in 7 post(s)
Thanks Martin, It's good to have results that correlate, somewhat.

Norm
Offline Jean Giraud  
#30 Posted : 16 September 2018 21:16:11(UTC)
Jean Giraud

Rank: Guest

Groups: Registered
Joined: 04/07/2015(UTC)
Posts: 6,868
Canada

Was thanked: 980 time(s) in 808 post(s)
Thanks for the Widget.

It ! compensates ! for not having more advanced solver.
It looks just a working trick for some data set.
It is not the conventional Weibull survival [failure is misleading].

The Hypergeometric.Edu is much different and standard as in "R"
There is one very important stat tool: Hypergeometric with and w/o replacement
RemToDo ... was under construction a while ago.

Statistics are pure convention, the matter is to agree with preferred convention

Jean

Stat Treasury_8 Weibull [log vraisemblance] Copy.sm (24kb) downloaded 10 time(s).
Stat Treasury_8 Weibull FailureAnalysis] Copy.sm (30kb) downloaded 11 time(s).
Weibull Widgets Copy.sm (60kb) downloaded 12 time(s).
Stat Hypergeometric Edu..sm (63kb) downloaded 10 time(s).
Users browsing this topic
2 Pages<12
Forum Jump  
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.