File Exchange

image thumbnail

System Identification using ANN

version 1.0.5 (2.15 KB) by Ayad Al-Rumaithi
System identification using artificial neural network example


Updated 09 Jul 2019

View License

This example file shows system identification using artificial neural network (ANN) of 2DOF system subjected to Gaussian white noise. The neural network consist of the following layers:

-Input layer: 2 nodes for the force at the current step and 2 nodes for the displacement at the previous step using open-loop feedback
-Hidden layer: 2 nodes for two inner states because there are 2 modes for 2DOF system
-Output layer: 2 nodes for the displacement

After training and getting the predicted output, the network was converted to closed-loop network and trained again (closed-loop networks uses predicted feedback from previous step instead of actual feedback). The predicted output from open-loop and closed-loop networks was compared with the actual output in a figure. It shows open-loop network is more accurate than closed-loop network due to the availability of actual output from the previous step.

Cite As

Ayad Al-Rumaithi (2020). System Identification using ANN (, MATLAB Central File Exchange. Retrieved .

Comments and Ratings (5)

Ayad Al-Rumaithi

I am using nonlinear autoregressive neural network with external input. The number of hidden layers is one.

Hakim Ahmad

may I know what type of ann you are using. i mean is it MLP or elman?


a good code

Ayad Al-Rumaithi

Dear Hakim,

I don’t have publications on this topic. But you can look at this paper
Derkevorkian, Armen, et al. "Nonlinear data‐driven computational models for response prediction and change detection." Structural Control and Health Monitoring 22.2 (2015): 273-288.

Hakim Ahmad

Assalamualaikum. a good code. do yo have any journal on this specific research?









Added closed-loop network


file comments

MATLAB Release Compatibility
Created with R2017b
Compatible with any release
Platform Compatibility
Windows macOS Linux