Answered
Testing different optimization methods on a state space model by solving ODEs
The closest infrastructure to this style of numerical optimization approach is "grey box identification". See idgrey, idnlgrey, ...

4 years ago | 0

| accepted

Answered
nlarx model initial conditions
You can prefix estimation data (both input and output signals) with nd zeros, where nd = maximum lag in the model. Initial condi...

4 years ago | 0

| accepted

Answered
How to force tfest to estimate the process with "only real poles" ?
TFEST cannot guarantee real poles. If you can work with <=3 poles and <=1 zero, try PROCEST. This is a process model estimator...

4 years ago | 3

| accepted

Answered
What is the difference between FRD and IDFRD
Both FRD and IDFRD are used to store Freqyency Response Data, that is, the complex frequency response vector (Mag.*exp(i*Phase))...

4 years ago | 1

| accepted

Answered
How to estimate the parameter in a customized transfer function
Grey-box identification is an option. You will need to write a function that takes K0 and a0 as inputs, and returns state-space ...

4 years ago | 0

Answered
Minimum input data resolution
Look up Nyquist Sampling Theorem. If you are sampling (hopefully with anti-aliasing) at 1Hz then you cannot theoretically captur...

4 years ago | 0

| accepted

Answered
Well-identified fitted process model does not behave like data on simulink
You are almost there. Convert the model into state-space form and use it for simulation. For initial conditions, you will need t...

4 years ago | 0

| accepted

Answered
How to calculate the transfer function for a 16 input system?
Try also TFEST. Although you might want to reduce the number of inputs by PCA or PLS analysis.

4 years ago | 0

Answered
nlarx model compare and predict (horizon kept 1) fit totally differs
The difference between (finite-horizon) prediction and simulation is a fundamental concept, something you could read books/artic...

4 years ago | 0

| accepted

Answered
Can nlgreyest() estimate open-loop unstable models?
With greyest, either parameterize K matrix using the ODE function, or choose to esitmate it separately by using the "Disturbance...

4 years ago | 0

Answered
System Identification of Closed Loop Data and Unstable Plant
The first reference: [1] System Identification — Theory For the User, Lennart Ljung, Section 13.4-13.5, 2nd ed, PTR Prentice Ha...

4 years ago | 0

| accepted

Answered
Accessing the GUI function programatically.
I will repeat Aditya Baru's comment as an answer. The App now supports MATLAB code generation (creating a function from the tas...

4 years ago | 1

| accepted

Answered
How to set parameters of Recursive Polynomial Model Estimator in Simulink
The Recursive Polynomial Model Estimator supports single output estimations only.

4 years ago | 0

| accepted

Answered
System Identification - Frequency Domain
Make an attempt with stability enforced. opt = tfestOptions('EnforceStability', true); model=tfest(f_data,6,opt) Also, you ...

4 years ago | 2

| accepted

Answered
System Identification Toolboox error dialog
What dataset are you using for validation? Does it suply the inputs and outputs that the model needs?

4 years ago | 0

Answered
Does the order (index) of inputs and outputs matter in MIMO system identification?
Yes the order matters since within a given noise level, there are many models that can explain the data. Settings related to sea...

4 years ago | 0

| accepted

Answered
System Identification Toolbox: How can we modify the starting parameters for the armax-algorithm?
You can set the A, B, C values explicitly, as in estimatedPolymodel.A = ARCoeff Or, call the IDPOLY constructor with A, B, C p...

4 years ago | 0

| accepted

Answered
How to identify a sytem by the System Identification Toolbox that is invertable ?
Tyically yes. If you are estimating state-space model, use "feedthough" name-value pair, as in ssest(Data, order, 'Feedthrough'...

4 years ago | 0

Answered
how to convert a xls to a data ensemble for import into diagnostic feature app
If the data is not too big to fit into MATLAB memory, I would suggest importing it into MATLAB first as a set of tables or timet...

4 years ago | 0

Answered
How can I find initial states for simulation?
Initial states show the effect of the environment on the system. They are not a property of the system to be determined uniquely...

4 years ago | 1

| accepted

Answered
System Identification toolbox: how to print estimated ARMA-coefficients for each iteration step in armax-algorithm
Use "full" display option, as in: opt=armaxOptions('Display','full'); estimatedPolymodel=armax(iddata(outputdata,inputdata,tsa...

4 years ago | 1

| accepted

Answered
how to use state space model?
If you want to reproduce the response of "predict" by (manual) simulation, you will need to generate the right prediction model ...

4 years ago | 1

Answered
how to plot on the same bode plot a manual function plot() with function bode()?
You could try: G = frd(f1(w),w); % assuming f1(w) is a complex numeric vector bode(G,f2)

4 years ago | 0

Answered
Residual analysis of 100% fit model using system identification toolbox
With simulated data with no noise, it is difficult to read the residual results since there is no baseline noise floor. That is,...

4 years ago | 0

| accepted

Answered
Fitting complex function to measurement values
Let X be your data matrix. [~,I]=unique(X(:,1),'stable'); h=X(I,2).*exp(1i*X(I,3)/180*pi); w=X(I,1)*2*pi; G=idfrd(h,w,'Ts',0...

4 years ago | 0

Answered
MISO system identification tool box step response
Yes, use LSIM with input: U = [u, u], where u is a step signal.

4 years ago | 0

| accepted

Answered
Merging MISO ARX models
What you need is horizontal concatenation, not MERGE which is about statistical merger of identical (same I/Os and model structu...

4 years ago | 0

| accepted

Answered
How to Improve the computation speed for 'ARMAX' function?
If you have access to Parallel Computing Toolbox, you could consider replacing the for-loop with a "parfor" loop. Other things t...

4 years ago | 0

Answered
How to estimate the inital state with an ssest output model
In short, you cannot. SSEST is a black-box identification function (unless you pass in a full initialized @idss model as input),...

4 years ago | 0

| accepted

Answered
Why cant I predict kstep ahead when adding System Identification models?
The issue is that algebra on identified models (plus, minus, series, parallel, feedback, inv etc) are not natively supported. Th...

4 years ago | 0

| accepted