Software developer, controls and identification products

Professional Interests: system identification, signal processing, controls

Answered

The system identification models I've exported to my work space do not match the results from the toolbox. Why is this happening?

What initial conditions are you using for simulation? Note that by default, the App estimates the initial conditions to maximize...

The system identification models I've exported to my work space do not match the results from the toolbox. Why is this happening?

What initial conditions are you using for simulation? Note that by default, the App estimates the initial conditions to maximize...

7 months ago | 0

| accepted

Submitted

Data based modeling of nonlinear dynamic systems using System Identification Toolbox

Perspectives on nonlinear identification using a throttle valve modeling example.

3 years ago | 10 downloads |

Answered

Transfer function from ARX-model

Does the example on http://www.mathworks.com/help/ident/ref/idnlarxlinearize.html help?

Transfer function from ARX-model

Does the example on http://www.mathworks.com/help/ident/ref/idnlarxlinearize.html help?

4 years ago | 0

Answered

Suggestion for System id toolbox

How about: NoiseModel = noise2meas(IdentifiedModel); Then use NoiseModel as a regular LTI model in Simulink and run it wi...

Suggestion for System id toolbox

How about: NoiseModel = noise2meas(IdentifiedModel); Then use NoiseModel as a regular LTI model in Simulink and run it wi...

5 years ago | 0

Answered

How do I model a time series in MATLAB using ARMA?

You should evaluate the quality of prediction results and perform some correlation analysis of the model residuals to verify how...

How do I model a time series in MATLAB using ARMA?

You should evaluate the quality of prediction results and perform some correlation analysis of the model residuals to verify how...

5 years ago | 0

Answered

Bump ---> Identification using invfreqs

Have you tried using the weight input argument? You can specify higher weights for the frequencies around which you are interest...

Bump ---> Identification using invfreqs

Have you tried using the weight input argument? You can specify higher weights for the frequencies around which you are interest...

6 years ago | 0

Answered

how to create a state-space model without disturbance and how does the disturbance influence the solution

The difference between the output of the model and the actual (measured) output is, in general, not going to be zero. So, there ...

how to create a state-space model without disturbance and how does the disturbance influence the solution

The difference between the output of the model and the actual (measured) output is, in general, not going to be zero. So, there ...

6 years ago | 1

| accepted

Answered

How to use the under knees code of state space model "ident" toolbox in mfile

Double click on estimated model icon. A window containing information on the model is shown. This window offers commands that wi...

How to use the under knees code of state space model "ident" toolbox in mfile

Double click on estimated model icon. A window containing information on the model is shown. This window offers commands that wi...

6 years ago | 0

| accepted

Answered

Use N4sid for system identification (for new data estimation)

Computing response of a model to a given input is called simulation. Use the idParametric/sim command for that as in: Ud = i...

Use N4sid for system identification (for new data estimation)

Computing response of a model to a given input is called simulation. Use the idParametric/sim command for that as in: Ud = i...

6 years ago | 0

| accepted

Answered

How to define a polynomial orders ARX model [na nb nk], for a MISO system

Look up the documentation of delayest on how to put some limits on the values to be tested. Delay can also be determined by a st...

How to define a polynomial orders ARX model [na nb nk], for a MISO system

Look up the documentation of delayest on how to put some limits on the values to be tested. Delay can also be determined by a st...

6 years ago | 0

Answered

improve the performance of system identification MISO

There are different types of choices to consider to get the most out of an identification: * Model orders: controlled by orde...

improve the performance of system identification MISO

There are different types of choices to consider to get the most out of an identification: * Model orders: controlled by orde...

6 years ago | 0

Answered

initial state in n4sid

Note that N4SID always estimates initial states when using time domain data; the InitialState option does not really help here. ...

initial state in n4sid

Note that N4SID always estimates initial states when using time domain data; the InitialState option does not really help here. ...

6 years ago | 0

Answered

Problem with ident GUI while predicting time-series

When predicting, the results depend upon the initial conditions. The GUI estimates the initial conditions to maximize the fit to...

Problem with ident GUI while predicting time-series

When predicting, the results depend upon the initial conditions. The GUI estimates the initial conditions to maximize the fit to...

6 years ago | 0

Answered

Prediction confidence intervals using a state-space model

For discrete-time systems, the forecasting model is same as the simulation (original) model. So you can use commands such as SIM...

Prediction confidence intervals using a state-space model

For discrete-time systems, the forecasting model is same as the simulation (original) model. So you can use commands such as SIM...

6 years ago | 1

| accepted

Answered

system identification command line- impulseest

The "60" refers to the length of the FIR filter. An FIR model typically has larger number of coefficients than a rational transf...

system identification command line- impulseest

The "60" refers to the length of the FIR filter. An FIR model typically has larger number of coefficients than a rational transf...

6 years ago | 0

| accepted

Answered

system identification gui - fit to estimation data

When the amx3322 model was estimated, it was estimated to minimize the 1-step ahead prediction error. The fit value between the ...

system identification gui - fit to estimation data

When the amx3322 model was estimated, it was estimated to minimize the 1-step ahead prediction error. The fit value between the ...

6 years ago | 1

Answered

How to design controller with control toolbox tools

You could try something along the lines of: m0 = idtf(ir, [1 zeros(1, length(ir)-1], 1/48000); m1 = balred(idss(m0), N); %...

How to design controller with control toolbox tools

You could try something along the lines of: m0 = idtf(ir, [1 zeros(1, length(ir)-1], 1/48000); m1 = balred(idss(m0), N); %...

6 years ago | 0

Answered

What is the criterion of best fit in system identification toolbox?

Normalized Root Mean Square (NRMSE) criterion is used to generate the fit % number. It is a measure of how much better the model...

What is the criterion of best fit in system identification toolbox?

Normalized Root Mean Square (NRMSE) criterion is used to generate the fit % number. It is a measure of how much better the model...

6 years ago | 1

Answered

My model output is different with use of Simulink and sim function

For iddata source, it helps to set Tstart to 0 since the simulation in Simulink starts there by default. That is, use: XX = idda...

My model output is different with use of Simulink and sim function

For iddata source, it helps to set Tstart to 0 since the simulation in Simulink starts there by default. That is, use: XX = idda...

6 years ago | 0

| accepted

Answered

how to use the model which is done by System identification toolbox?

Did you set the initial conditions correctly? See: http://www.mathworks.com/matlabcentral/answers/72338-how-fix-the-problem-...

how to use the model which is done by System identification toolbox?

Did you set the initial conditions correctly? See: http://www.mathworks.com/matlabcentral/answers/72338-how-fix-the-problem-...

6 years ago | 0

Answered

how to find parameters (A and B ) for a non linear arx model and how can i convert this non linear arx model into transfer function

"Converting" a nonlinear model into a linear one is an exercise in linearization. See idnlarx/linearize and idnlarx/linapp funct...

how to find parameters (A and B ) for a non linear arx model and how can i convert this non linear arx model into transfer function

"Converting" a nonlinear model into a linear one is an exercise in linearization. See idnlarx/linearize and idnlarx/linapp funct...

6 years ago | 0

Answered

About the underlying principle of the function "predict"

The prediction is based on computing a kalman filter. See chapter 3 in "System Identification, Theory for the User", 2nd ed, Len...

About the underlying principle of the function "predict"

The prediction is based on computing a kalman filter. See chapter 3 in "System Identification, Theory for the User", 2nd ed, Len...

6 years ago | 0

Answered

Model generated by system identification tool when transported into simulink, does not match and is unstable

Hi Saurabh, Could you try running the SIM command in the MATLAB Command Window (e.g. y = sim(pss3, input)) and check if the sim...

Model generated by system identification tool when transported into simulink, does not match and is unstable

Hi Saurabh, Could you try running the SIM command in the MATLAB Command Window (e.g. y = sim(pss3, input)) and check if the sim...

6 years ago | 0

Answered

Is there a way to output the final simulated time-response from procest()?

Ah, I see it now: it is the InputOffset that you estimated or supplied during process model estimation but perhaps did not speci...

Is there a way to output the final simulated time-response from procest()?

Ah, I see it now: it is the InputOffset that you estimated or supplied during process model estimation but perhaps did not speci...

6 years ago | 0

| accepted

Answered

Bode diagram of poor input/output signal

bode(sys,[w1 w2],'.'), where w1 and w2 are your two input frequencies and sys is the identified model.

Bode diagram of poor input/output signal

bode(sys,[w1 w2],'.'), where w1 and w2 are your two input frequencies and sys is the identified model.

6 years ago | 0

| accepted

Answered

Identification of MISO / MIMO models based on transfer functions: poles multiply contained in resulting model

The repeated poles are probably a result of concatenation; do pole(ss(zpk(rss(2,2,2)))) to check this phenomenon. You can use SM...

Identification of MISO / MIMO models based on transfer functions: poles multiply contained in resulting model

The repeated poles are probably a result of concatenation; do pole(ss(zpk(rss(2,2,2)))) to check this phenomenon. You can use SM...

6 years ago | 0

Answered

Is there a way to output the final simulated time-response from procest()?

Hi James, If the estimation focus is "simulation", you should not really see this difference. What MATLAB release are you seein...

Is there a way to output the final simulated time-response from procest()?

Hi James, If the estimation focus is "simulation", you should not really see this difference. What MATLAB release are you seein...

6 years ago | 0

Answered

How to fit data into nonlinear model

Is there an input/output relationship among the 2 variables? That is, does changing one affects the other in a cause-effect mann...

How to fit data into nonlinear model

Is there an input/output relationship among the 2 variables? That is, does changing one affects the other in a cause-effect mann...

6 years ago | 0

Answered

What are the 7 states for the model?

Can you elaborate on what you mean by "defining the states"? If you want to calculate the state-trajectory (as a function of tim...

What are the 7 states for the model?

Can you elaborate on what you mean by "defining the states"? If you want to calculate the state-trajectory (as a function of tim...

6 years ago | 0

Answered

What is the cost function in the optimization problem of the pem and greyest functions for multiple outputs?

The weighting matrix by default is inverse of noise variance which is simultaneously estimated along with the model parameters. ...

What is the cost function in the optimization problem of the pem and greyest functions for multiple outputs?

The weighting matrix by default is inverse of noise variance which is simultaneously estimated along with the model parameters. ...

6 years ago | 0