System impulse response and Convolution by matlab

Hi everyone, i am very new to matlab, and would like to konw how to obtain y(t) from below

Generate a unit step function as the input function, x(t), and an exponentially decay function as the impulse response function, h(t), such as h(t)=exp(-t/2) (note: 2 is the time constant of the system dynamic response). Using MATLAB to calculate the output of the system, y(t).

Thank you so much in advance

[Seikh Rana]

Convolution (random position zero) in matlab plz help me

[Seikh Rana]

</matlabcentral/answers/uploaded_files/114156/convooooo.JPG> this picture problem solved in matlab plz help me

Answer 1

[psyx21]

hey friends what will be the peak acceleration response for 100g 6ms half sine pulse with zeta=0. I need a matlab script with natural frequency on x axis..thanks

Answer 2

[Paulo Silva]

t=0:0.1:10;
u=0*t;
u(t>=0)=1;
h=exp(-t/2);
y=u.*h;
plot(t,y)

[gedaa]

thank you sooo much, but I have one more question if it's not 2 much, what is u=0*t and how did u come up with that?

[Paulo Silva]

it's the same as

u=zeros(size(t))

It creates a variable with the same size as t but full of zeros. I learned that trick here at matlabcentral :)

[David Young]

Paulo - I just wonder if this is right. In the time domain, don't you need to convolve the input and the impulse response rather than multiplying them?

[Paulo Silva]

Yes you are correct, try

t=0:0.1:10;

u=0*t;

u(t>=0)=1;

h=exp(-t/2);

y=conv(u,h);

plot(y)

[Paulo Silva]

Correction for the amplitude and time scale

T=0.1;

t=0:T:10;

u=0*t;

u(t>=0)=1;

h=exp(-t/2);

y=conv(u,h);

plot(t,T*y(1:numel(t)))

[gedaa]

ok, so if the problem asks for a pulse function instead of a unit step function, it would be the exact same codes except with the addition of PulseWidth=1, am i correct?

[Paulo Silva]

I'm not sure about the amplitude at

plot(t,T*y(1:numel(t)))

with the step function is good compared to

step(tf([1],[1 1/2])) but with the impulse it's plot(t,y(1:numel(t))) without the T, I can't figure out why that happens.

[gedaa]

hey, paulo, i actually checked with my professor and he said that T doesn't need to be multiplied in the last line. Thank you soo much

Answer 3

[Arpan Patel]

t=0:0.1:10;

u=0*t;

u(t>=0)=1;

h=exp(-t/2);

y=u.*h;

plot(t,y)