-23023 error in Non linear system solver

[ticofs]

Hi to all !!.

Can someone help me  with this  problem.

I am trying to solve a linaer system with 5 variables but the solver gives me the -23023 error. These are the equations:

((ln(9,616678E+0/c)-ln(9,616678E+0/3,854417E-9))*25,774472E-3)/(-1*a)+9,616678E+0*(d/c)*exp(1/n)+25,774472E-3*(1/b)*ln(9,616678E+0/c)

((ln(6,700912E+0/c)-ln(6,700912E+0/4,069466E-9))*25,804633E-3)/(-1*a)+6,700912E+0*(d/c)*exp(1/n)+25,804633E-3*(1/b)*ln(6,700912E+0/c)

((ln(6,139159E+0/c)-ln(6,139159E+0/3,854417E-9))*25,774472E-3)/(-1*a)+6,139159E+0*(d/c)*exp(1/n)+25,774472E-3*(1/b)*ln(6,139159E+0/c)

((ln(3,778458E+0/c)-ln(3,778458E+0/3,566005E-9))*25,731386E-3)/(-1*a)+3,778458E+0*(d/c)*exp(1/n)+25,731386E-3*(1/b)*ln(3,778458E+0/c)

((ln(2,514514E+0/c)-ln(2,514514E+0/3,147102E-9))*25,662447E-3)/(-1*a)+2,514514E+0*(d/c)*exp(1/n)+25,662447E-3*(1/b)*ln(2,514514E+0/c)

I also try to write the equation without the technical notation but I have the same error.

I don´t find any post for this error.

Thank you.

[sendia]

Hi ticofs,

The error is something related with variable initialization. Can you post a code example with your implementation?

[ticofs]

Here is the code.

Thank you for the help.

[johnsold]

You are using commas for decimal separators.  I think the formual parsers do not allow that.  Try replacing the commas with periods.

Lynn

[sendia]

Hi ticofs,

I had the same error if the START and END inputs are not initialized. Initialize both inputs

[ticofs]

I replace the commas with periods and now I have the -23001 error, anyone knows why?.

Thank you

[ticofs]

I try to initililize start and end inputs, but this cause not effect in the program.

Thank you for the idea.

[johnsold]

What data values do you use for Start and End?

Lynn

[johnsold]

I tried changing the numeric format. I got rid of the exponential notation and it seemed to eliminate the errors.  Then I restored the original format and the errors are still gone????

I notice that the output on Array 5 is quite different from run to run.  Is your system of equations such that it may be extremely sensitive to numerical conditions like round off errors?  Is it close to being singular?  I used all zeros for Start and all ones for End.

You are using a quite old version of LV.  When I open your VI in LV 2011 it warns that the solver VI is obsolete.  I connected your equations to the newer version.  It produces the same errors when there are errors but I never get any output on f(Zeros).

Lynn

[ticofs]

Hi!.

The program takes the values from the intensity of the ccd of one camera, and with this information and the temperature, we must to calculate the other five variables. We take 5 pictures with different conditions to have 5 equations to solve the system. 

The values for start and end should be between these values:

       Start      End

a     0,1         100

b     1            10E8

n     0,5         3

d     1E-13     10000E-11

c     0,08       2

I rid the exponential notation and used zeros in the start and ones in the end  but I still have the same error, you know why?. Can you send me your vi?. I send you what I have done.

I know that it´s going to be difficult to have solutions from these system of equations. As you say, the system of equation is extremely sensitive to numerical conditions and I also think that it could be quite near to be singular. Now, I only want to eliminate the error to experiment with different values.

The version is the 6.1. I can´t change the version because the experiment code is programing in this version.

Thank you.