Hi,
I'm working on applying a nonlinear error model on a dataset. the model is for part of the data and the model is applied on the rest of the data. when i applied the model inside the loop, the results diverge while when i apply it outside the loop the results are converge. this was fro the same dataset.
I think it must be converge when you apply inside the loop, and diverge when you applied outside thee loop .you have some mistake give more details about boundary condition ,signal characteristics,wire connecting and others .
Thanks Jumaa for your reply,
the issue is i'm applying a moving window based non-linear error modelling for a data of a sensor. the first part of the data is used to create the model. afterward, the rest data points are used to implement the model. i created the model and to apply it i have to apply it in a for loop because of the windowing issue. therefore, i'm looking for a solution for the delays effect.
as you know the difference equation is depending one the previous states of the input and the output making IIR filter. the delays in these states are not mitigated in the loop. when i apply the model directly to the data without windowing, the data are converging to the reference.
- Badges
- Science topic