Data structure for system identification toolbox for multiple input multiple output data

Hello,

I'm trying to analyze data in the system identification toolbox.

The imput variables are time, pressure, angle and distance and the outputs are temperature values from T1 to T6, like in the attached data sheet.

I'm struggeling to get the data into the toolbox.

Can I load the whole data set into the toolbox?

This is what I tryied so far:

clc, close all, clear all
data = xlsread('MiMO.xlsx');
save MiMO.mat data;
load MiMO.mat

time=data(:,1);
pressure=data(:,2);
angle=data(:,3);
distance=data(:,4);
T1=data(:,5);
T2=data(:,6);
T3=data(:,7);
T4=data(:,8);
T5=data(:,9);

ST=1; % Sampling time in seconds

MiMO_Data=iddata(['time';'pressure';'angle','distance'],['T1','T2','T3','T4','T5'],ST);

Error using vertcat
Dimensions of arrays being concatenated are not consistent.

MiMO_Data.ImputName={'time';'pressure';'angle','distance'};
MiMO_Data.OutputName={'T1','T2','T3','T4','T5'};

systemIdentification;

I used iddata structure to get the whole data into the system identification toolbox,

but the structure seems not right.

[Mathieu NOE]

hello

before we even start with how to speak matlab code, simply look at the data :

pressure, angle and distance are constant all the time so you can't use that as input variables (as they don't vary )

also even if they vary we should ask ourself if they are input or output variables

basically , all yout T values just vary linearly with time but we don't have the "physical" input variable in your data

it could be that there is only one single independant input variable in your problem so it's basically even simpler to modelize.

[Steffen B.]

Hello Mathieu,

thanks for your answer.

The data from the table is only exemplary and not relevant. They are only used to develop the method with which the real values are then analysed. They are just placeholders for the real data that I don't have yet. But they will have multiple input and output variables.

[Mathieu NOE]

hello again

this line

MiMO_Data=iddata(['time';'pressure';'angle','distance'],['T1','T2','T3','T4','T5'],ST);

is wrong because instead of giving data to iddata, you're giving char arrays (with size error so that's why you get : Error using vertcat)

try this instead :

data = xlsread('MiMO.xlsx');

ST=1; % Sampling time in seconds
MiMO_Data=iddata(data(:,1:4),data(:,5:9),ST); % MiMO_Data=iddata(['time';'pressure';'angle','distance'],['T1','T2','T3','T4','T5'],ST);
whos

  Name            Size             Bytes  Class     Attributes

  MiMO_Data      21x4x5             8247  iddata              
  ST              1x1                  8  double              
  ans             1x33                66  char                
  data           21x9               1512  double              

[Steffen B.]

Hello Mathieu,

thanks for your support.

Now it's working just fine.

Have a nice day.

[Mathieu NOE]

hello Steffen

glad I could help

have a nice day too