Why do I get 'Undefined function or variable'?

I'm attempting to run 'ode45' and changed the function file code to include an input vector 'u'. Now I get an error stating that 'u' is undefined. I don't know where to define 'u', since it is defined in the function?

(Undefined function or variable 'u'.)

(Error in ode45_ex>@(t,x)DOF1damped(t,x,u,w,u0,m,k,c,varargin))

the function file is:

function [dx, y] = DOF1damped(t, x, u, w, u0, m, k, c, varargin)
    u = [u0*sin(w*t); u0*w*cos(w*t)];
    y = [x(1);x(2)];
    dx = [x(2) ;((-k/m).*x(1) - (c/m).*x(2) + (k/m).*u(1) + (c/m).*u(2))];
end

the ode file is:

clear 
    close all
    clc

    u0 = 1; %amplitude ()
    w = 1; %hz
    m = 0.85048569375; %kg
    k  = 356.9085; % N/m
    c = 10; %N/m*s

    tspan = [0 10];
    x0 = [0;0];
    [t, x] = ode45(@(t,x)DOF1damped(t, x, u, w, u0, m, k, c, varargin), tspan, x0); 

    data = iddata(x, t, .1);

    plot(t, x);

Thanks for any help

Answer 1

[Stephan]

If u is calculated inside the function it should not be an input argument of the function. Also delete varargin - no need - results in error, because it is a matlab function already:

u0 = 1; %amplitude ()
w = 1; %hz
m = 0.85048569375; %kg
k  = 356.9085; % N/m
c = 10; %N/m*s
tspan = [0 10];
x0 = [0;0];
[t, x] = ode45(@(t,x)DOF1damped(t, x, w, u0, m, k, c), tspan, x0);
data = iddata(x, t, .1);
plot(t, x);

function [dx, y] = DOF1damped(t, x, w, u0, m, k, c)
u = [u0*sin(w*t); u0*w*cos(w*t)];
y = [x(1);x(2)];
dx = [x(2) ;((-k/m).*x(1) - (c/m).*x(2) + (k/m).*u(1) + (c/m).*u(2))];
end

[Ben25]

Thanks Stephan, this works perfectly in order to plot the data from the ode file, however it creates a new problem. These two files are used as sample data in an idnlgrey function:

filename = 'DOF1damped';
Order = [2 0 2];

u0 = 1; %amplitude ()
w = 1; %hz
m = 0.85048569375; %kg
k = 356.9085; % N/m
c = 10; %N/m*s

Parameters = {u0,w,m,k,c};
InitialStates = [0; 0];
Ts = 0;


nlgr = idnlgrey(filename, Order, Parameters, InitialStates, Ts, 'Name','DOF1grey');

nlgr = setinit(nlgr, 'Fixed', {false false}); 
% Estimate the initial states.
setpar(nlgr,'Fixed',{true true true true false});
opt = nlgreyestOptions('Display', 'on');

nlgr = nlgreyest(data,nlgr,'Display','Full');
c_est = nlgr.Parameters(5).Value;
[nonlinear_c_est, dnonlinear_c_est] = getpvec(nlgr,'free')


compare(data,nlgr)

I updated the order to include no input, and I received this error.

Error using iddata/nlgreyest (line 59)
The number of inputs and outputs of the model must match that of the data.

Error in DOF1grey (line 21)
nlgr = nlgreyest(data,nlgr,'Display','Full');
 

[Stephan]

should not oder be:

Ordner = [1 0 2]

[Ben25]

This doesn't work either, I believe the first number should still be 2, as it represents the y vector, which has an order of 2. I still do not know where the problem is, as there are definitely 2 outputs and 2 states.

[Ben25]

My problem is very similar to the section 'Perform Nonlinear Grey Box Estimation' on this example https://ww2.mathworks.cn/help/ident/ug/estimating-coefficients-of-odes-to-fit-given-solution.html#d117e41137 . I don't see the problem.

[Stephan]

Maybe it helps if you attachyour data