Hello I am currently working with Non Linear Models and the System Ident. Toolbox in my university. I managed to get a good fit for the data and my model,
however I am not sure how to interpret the Autocorrelation of the residuals for my model, which is shown below. They seem to be okay except for the range of
-10/+10 samples. Do you have any idea how to interpret this results and how I can try to change them? I tried different orders but nothing really changed, just the shape
seemed to move a bit.
I thought it might be because of the data itself. I am measuring temperatures which have a an approx error of -4/+4 C° because of the sensor. Could this be a reason? Also what exactly does this graph mean? I interpret it like this: When looking at big sample quantities of +10 and more my mode is doing fine, however if i look in the near vicinity of my data, lets say sample 293 and its -5/+5 closest neigbours there is a relatively big error in measuring.