Peak discovery of gauged signal

We make use of an information procurement card to take analyses from a tool that raises its signal to an optimal and afterwards drops back to near the initial value. To locate the peak value we presently look the array for the highest possible analysis and also make use of the index to establish the timing of the peak value which is made use of in our estimations.

This functions well if the highest possible value is the optimal we are seeking yet if the tool is not functioning appropriately we can see a 2nd optimal which can be greater than the first optimal. We take 10 analyses a 2nd from 16 tools over a 90 2nd duration.

My first ideas are to cycle via the analyses examining to see if the previous and also next factors are much less than the existing to locate an optimal and also construct an array of optimals. Possibly we need to be considering a standard of a variety of factors either side of the existing placement to permit sound in the system. Is this the most effective means to continue or exist far better strategies?


We do make use of LabVIEW and also I have actually examined the LAVA forums and also there are a variety of intriguing instances. This becomes part of our examination software program and also we are attempting to stay clear of making use of way too many non - typical VI collections so I was wishing for responses on the process/algorithms entailed as opposed to details code.

0
2019-05-09 11:13:46
Source Share
Answers: 4

I assume you intend to cross-correlate your signal with an anticipated, prototype signal. Yet, it has actually been such a long period of time given that I researched signal handling and also also after that I really did not take much notification.

0
2019-05-12 02:57:51
Source

I do not recognize significantly concerning instrumentation, so this could be entirely not practical, however it could be a handy various instructions. If you recognize just how the analyses can fall short, and also there is a particular period in between optimals offered such failings, why refrain slope descent at each period. If the descent brings you back to a location you've looked in the past, you can desert it. Relying on the form of the experienced surface area, this additionally could aid you locate optimals faster than search.

0
2019-05-12 02:56:09
Source

You can attempt signal averaging, i.e. for each and every factor, balance the value with the bordering 3 or even more factors. If the sound spots are massive, after that also this might not aid.

I become aware that this was language agnostic, yet presuming that you are making use of LabView, there are great deals of pre - packaged signal handling VIs that feature LabView that you can make use of to do smoothing and also sound decrease. The NI forums are a wonderful area to get even more specialized aid on this type of point.

0
2019-05-10 06:06:46
Source

You can use some Standard Deviation to your reasoning and also notice optimals over x%.

0
2019-05-10 05:41:58
Source