AI Assistant
Notifications
Clear all

Repeatability of a Survey with RTK

47 Posts
18 Users
0 Reactions
1,714 Views
Kent McMillan
(@kent-mcmillan)
Posts: 11416
Member
Topic starter
Translate
English
Spanish
French
German
Italian
Portuguese
Russian
Chinese
Japanese
Korean
Arabic
Hindi
Dutch
Polish
Turkish
Vietnamese
Thai
Swedish
Danish
Finnish
Norwegian
Czech
Hungarian
Romanian
Greek
Hebrew
Indonesian
Malay
Ukrainian
Bulgarian
Croatian
Slovak
Slovenian
Serbian
Lithuanian
Latvian
Estonian
 

> Get some hands on exprieence and frame the conversation as how-to-improve instead of how to slam or just win and argument, and they maybe more pople will be ready to play.

Okay, I'll mark you down as not interested in the topic from the standpoint of error analysis and we'll carry on. The idea that we can't answer the problem as posed without holding a network RTK rover is ridiculous on its face, as will be seen.


 
Posted : August 23, 2012 6:49 pm
ashton
(@ashton)
Posts: 566
Member
Translate
English
Spanish
French
German
Italian
Portuguese
Russian
Chinese
Japanese
Korean
Arabic
Hindi
Dutch
Polish
Turkish
Vietnamese
Thai
Swedish
Danish
Finnish
Norwegian
Czech
Hungarian
Romanian
Greek
Hebrew
Indonesian
Malay
Ukrainian
Bulgarian
Croatian
Slovak
Slovenian
Serbian
Lithuanian
Latvian
Estonian
 

OK, I recognize "expected value" as a probability term of art with a definite definition. For nicely shaped distributions, it would coincide with the peak of the distributions. "Most probable value" is not a term of art that I recall or can find in handy texts; unless it is a term used in textbooks I haven't encountered, one could quibble about what it means.


 
Posted : August 23, 2012 7:26 pm
Kent McMillan
(@kent-mcmillan)
Posts: 11416
Member
Topic starter
Translate
English
Spanish
French
German
Italian
Portuguese
Russian
Chinese
Japanese
Korean
Arabic
Hindi
Dutch
Polish
Turkish
Vietnamese
Thai
Swedish
Danish
Finnish
Norwegian
Czech
Hungarian
Romanian
Greek
Hebrew
Indonesian
Malay
Ukrainian
Bulgarian
Croatian
Slovak
Slovenian
Serbian
Lithuanian
Latvian
Estonian
 

As a practical matter, 0.003 ft., (nominally 1mm) is the least significant unit of distance for nearly all purposes of land surveying (meaning boundary surveying). A distance such as 125.31415978 ft. has no meaning beyond possibly the first three places in the decimal, so there's not much point in dealing with a continuous probability distribution anyway. A discrete distribution with 0.003 ft. steps is for all practical purposes more than sufficient.

I made the calculation using 1,000 trials from random values with a Gaussian distribution (mean = 0, sigma = 0.02 ft.) and used the results to construct a discrete distribution with 0.003 ft. steps.


 
Posted : August 23, 2012 7:50 pm
Brooks Cooper
(@brooks-cooper)
Posts: 65
Member
Translate
English
Spanish
French
German
Italian
Portuguese
Russian
Chinese
Japanese
Korean
Arabic
Hindi
Dutch
Polish
Turkish
Vietnamese
Thai
Swedish
Danish
Finnish
Norwegian
Czech
Hungarian
Romanian
Greek
Hebrew
Indonesian
Malay
Ukrainian
Bulgarian
Croatian
Slovak
Slovenian
Serbian
Lithuanian
Latvian
Estonian
 

How do you know you've never got the same bad reading twice?


 
Posted : August 23, 2012 9:26 pm
Brooks Cooper
(@brooks-cooper)
Posts: 65
Member
Translate
English
Spanish
French
German
Italian
Portuguese
Russian
Chinese
Japanese
Korean
Arabic
Hindi
Dutch
Polish
Turkish
Vietnamese
Thai
Swedish
Danish
Finnish
Norwegian
Czech
Hungarian
Romanian
Greek
Hebrew
Indonesian
Malay
Ukrainian
Bulgarian
Croatian
Slovak
Slovenian
Serbian
Lithuanian
Latvian
Estonian
 

What is you email Mr. Schrock? I have something for you.


 
Posted : August 23, 2012 9:35 pm

Sat Al
(@sat-al)
Posts: 197
Member
Translate
English
Spanish
French
German
Italian
Portuguese
Russian
Chinese
Japanese
Korean
Arabic
Hindi
Dutch
Polish
Turkish
Vietnamese
Thai
Swedish
Danish
Finnish
Norwegian
Czech
Hungarian
Romanian
Greek
Hebrew
Indonesian
Malay
Ukrainian
Bulgarian
Croatian
Slovak
Slovenian
Serbian
Lithuanian
Latvian
Estonian
 

You can do it in a lab, but the possibility of doing it, without intention, in the field in reasonable GPS conditions is, in my opinion, not a part of this discussion.

> The way ambiguity fixing works, and the way it has to run through thousands of candidate resolutions, that would be statistically so unlikely as to bordre on near impossible. Not impossible, but remember, there are multiple epochs analyzed for each stored "shot". So one shot is really a group of shots ,so you''d ahve to get the same bad shot 6, 8 , 10, 120, 240 etc or more times depending on how you did it.


 
Posted : August 23, 2012 9:54 pm
Kris Morgan
(@kris-morgan)
Posts: 3855
Member
Translate
English
Spanish
French
German
Italian
Portuguese
Russian
Chinese
Japanese
Korean
Arabic
Hindi
Dutch
Polish
Turkish
Vietnamese
Thai
Swedish
Danish
Finnish
Norwegian
Czech
Hungarian
Romanian
Greek
Hebrew
Indonesian
Malay
Ukrainian
Bulgarian
Croatian
Slovak
Slovenian
Serbian
Lithuanian
Latvian
Estonian
 

Kent

No, the thrust of what I'm saying is that the math problem is pigeon holed and doesn't reflect the total, and while I can't express the total in mathematical terms, I can qualify it with experience.

As I said earlier, your math is spot on, and I agree with the answers you're getting.


 
Posted : August 24, 2012 6:22 am
Page 3 / 3