Guessing numbers

**Roadhog03** · 06-05-2014, 10:39 AM

I have a table some numbers are filled in others are not I need to find a way to guess at what the other numbers could be.
the top row and left side are labels throttle position on top and RMP on bottom. I need to guess at what the zeros are based on the numbers that are there.

Rear 0 2 5 7 10 15 20 25 30 40 60 80 100
500 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
750 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
1000 11.8 0.0 -12.2 -6.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
1125 -0.1 8.2 -15.1 -9.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
1250 -43.4 5.6 24.7 -7.6 -0.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
1500 -54.0 6.5 4.0 0.0 -1.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
1750 -54.5 15.1 6.2 19.4 16.1 0.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0
2000 -54.5 19.2 13.8 12.3 12.6 8.9 4.9 0.0 0.0 0.0 0.0 0.0 0.0
2250 7.1 10.0 16.9 16.6 17.7 8.8 3.5 -0.7 0.0 0.0 0.0 0.0 0.0
2500 17.2 3.4 15.6 15.8 15.1 8.9 4.9 4.5 0.0 5.8 0.0 0.0 0.0
2750 10.8 5.4 12.6 13.1 12.4 8.0 4.7 6.5 8.8 4.3 0.0 0.0 0.0
3000 13.5 0.0 6.8 0.0 15.3 13.7 11.4 19.2 18.3 7.7 0.0 0.0 0.0
3500 4.1 0.0 0.0 10.8 0.0 0.0 3.5 15.1 7.2 7.0 0.0 0.0 0.0
4000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
4500 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
5000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
5500 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
6000 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
6500 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

**MrShorty** · 06-05-2014, 10:58 AM

This kind of thing is usually a regression/curve fitting kind of problem. You have a function z=f(RPM, TP). This kind of algorithm looks something like:

1) Select an expression for f that represents the process being described. Do you know what kind of function f() should be? I plotted the data up and looked at it quickly, and I have no idea what kind of "simple" function would explain the variations you have. This is often the hardest part of this kind of problem.
2) Using a suitable regression algorithm (LINEST() or Solver in Excel), regression parameters for f from the given data.
3) Solve for the unkown z's using the given RPM and TP and f as regressed in step 2.

You did not say what z represents, and I don't know my way around these kind of engine calculations, so I'm not sure if I will be much help in selecting f. However, once you have some idea of what f should look like, we should be able to help you program a spreadsheet to perform the regression and compute the missing values.

**Roadhog03** · 06-05-2014, 11:42 AM

Lets try this a different way maybe.

The table below shows the data as it was corrected what you are looking at is the measure of air fuel at a given throttle present and RPM that give a Air Fuel ratio of 14.6 problem is only some cells are reached with the OEM o2 sensors I am thinking there might be a way to mathematical guess at what the others could be based on the ones we have.

in the attachment the green area are confirmed good data now I want to make a good guess as to what the other areas could be

I am making sense,

Jug 1.png

0 2 5 7 10 15 20 25 30 40 60 80 100
500 70.0 73.0 77.0 78.0 78.0 76.0 72.0 66.0 63.0 63.0 63.0 63.0 63.0
750 70.0 73.0 77.0 78.0 78.0 76.0 72.0 66.0 63.0 63.0 63.0 63.0 63.0
1000 59.6 72.0 72.5 70.3 80.0 79.0 75.0 71.0 69.0 69.0 69.0 69.0 69.0
1125 60.2 68.6 65.2 64.3 82.0 81.0 79.0 75.0 74.0 74.0 74.0 74.0 74.0
1250 54.5 62.8 79.6 69.5 71.4 85.0 83.0 81.0 80.5 81.0 83.0 85.0 85.0
1500 50.2 66.7 65.9 74.7 76.2 93.0 94.0 93.0 91.0 91.0 95.0 98.0 100.0
1750 47.6 58.1 64.5 78.1 75.4 91.0 97.0 93.0 90.0 90.0 93.0 95.0 96.0
2000 49.5 59.7 67.4 75.2 81.1 82.9 88.3 92.0 89.0 85.0 86.0 86.0 86.0
2250 60.2 61.6 67.7 78.7 87.4 95.7 94.1 95.0 90.0 87.0 86.0 85.0 85.0
2500 61.8 73.3 73.3 78.9 86.9 93.6 94.1 99.3 91.5 78.6 89.0 88.0 88.0
2750 66.7 70.3 67.8 82.8 89.6 97.2 95.7 97.0 92.2 84.2 93.0 93.0 93.0
3000 57.6 75.0 77.9 91.0 88.2 97.4 93.2 85.8 95.9 92.6 96.0 96.0 96.0
3500 67.5 77.5 86.0 81.5 103.0 113.0 112.3 108.0 104.9 101.0 105.0 104.0 104.0
4000 68.0 76.5 85.0 94.0 102.5 113.0 116.0 114.0 112.0 109.0 106.0 105.0 105.0
4500 63.0 69.0 75.0 82.0 90.0 105.0 109.0 108.0 107.0 105.0 104.0 104.0 104.0
5000 61.0 67.0 73.0 79.0 88.0 104.0 108.0 106.0 105.0 103.0 100.0 98.0 97.0
5500 61.0 67.0 73.0 79.0 88.0 104.0 107.0 106.0 105.0 103.0 99.0 97.0 96.0
6000 61.0 67.0 73.0 79.0 88.0 104.0 107.0 106.0 105.0 103.0 98.0 96.0 95.0
6500 61.0 67.0 73.0 79.0 88.0 104.0 107.0 106.0 105.0 103.0 98.0 96.0 95.0

**MrShorty** · 06-05-2014, 12:23 PM

It still looks like a curve fitting/regression type problem. As I indicated earlier, I do not readily see a function that will fit the data.

Perhaps if we could look at this from a theoretical perpective. I'm not any kind of automotive/mechanical engineer, so I'm mostly speculating here. Assuming a 4 stroke engine and that we know the displacement, I believe we can readily say that an engine processes a volume d (d=displacement) of air + fuel every 2 rotations. From the given engine speed (RPM), we should be able to calculate how much air+fuel passes through the engine at a given engine speed. It has never been clear to me what the units are on that 14.6 ratio (mass/mass, mole/mole, vol/vol, other??). From the known engine speed and discplacement, and assuming values for manifold temperature and pressure, we should be able to calculate how much air and fuel the engine would theoretically use. By figuring out how manifold pressure and temperature are related to engine rpm and throttle position, we could further estimate how these parameters figure into the quantity we are calculating.

Recognizing that this procedure is probably grossly oversimplified, perhaps it will give us some sense of what form we might use for f in our regression equation.

Can I ask -- what is your background? Are you an engineer (or engineer in training) who would know the equations that are used in describing engine operation, or are you a hobbyist trying to understand better how an engine works?

**Roadhog03** · 06-05-2014, 12:35 PM

Hobbyist just trying to learn to tune my EFI Harley

**MrShorty** · 06-05-2014, 01:08 PM

As hobbyists, it can be real hard to mathematically understand engines, because we do not have the training in math and engine theory to really model these things mathematically.

It is still unclear to me what z represents. Air flow (mass air reading)? O2 sensor reading? something else? obviously you are trying to optimize the A/F ratio, but I'm still not clear on how this is part of calculating A/F ratio.

One observation: Above about 7-10% throttle, the "known good" values and the complete curves kind of cluster together. If we are going to do this "empirically" I might be tempted to fit the data above 10% to one equation and the data below 10% to another equation.

Looking at it differently, if I plot z against throttle position, it looks like engine speed may not even be important in determining z -- perhaps fit the entire data set as z=f(TP) and ignore engine speed.

Guessing numbers

LinkBack

Thread Tools

Rate This Thread

Display

Guessing numbers

Re: Guessing numbers

Re: Guessing numbers

Re: Guessing numbers

Re: Guessing numbers

Re: Guessing numbers

Thread Information

Users Browsing this Thread

Similar Threads

Detect numbers in a cell range and then change cells based on what numbers are found

[SOLVED] Convert all numbers stored as text or custom formatted to numbers &no decimals - 40 sheets

Editing Social Security Numbers from current numbers to some with XXXX inserted automatica

[SOLVED] How to merge a numbers from 3 cells, eliminate repetitive numbers, and sort such numbers?

Filtering (I'm guessing)

Bookmarks

Bookmarks

Posting Permissions