How can I determine the time t when a specific value y is reached in SIMULINK?

Hi everyone,

I'm working on a simulation on SIMULINK which outputs me a value y(t).

I now want to know at what time y(t) reaches a specific value y_1.

So I need to determine t(y_1) = t_1 and somehow display it in SIMULINK?

Also I want to used that value t_1 for further calculations.

Is there any block that I can use for that?

Best regards

Enea

Answer 1

[Timo Dietz]

There are multiple ways of doing this. One is to attach the source block "clock" to a triggered subsystem which contains just a connection between in and out. The trigger input is fed by a comparator which checks for y(t) = y_1.

The triggered block now takes over the simulation time input to its output only in case the comparator triggers.

[Enea Baumann]

Hi there,

thank you very much for your answer!

That worked perfectly!

One more question. I now have as a output of the triggered subystem a stepfunction that reaches t_1 at t_1.

How can I now convert the final value (t_1) of that output into a normal constant block?

[Timo Dietz]

Hm, would help if you could describe in more detail for what exactly you want to use that signal. It actually is constant - after t_1.

[Enea Baumann]

Okay, so the problem I have now is the following.

I basically have to equations to solve with SIMULINK.

First I want to solve equation 1. When equation 1 is solved I want to determine the time t_1 when the solution of euqation 1 reaches y_1. This worked so far.

Now I want to use t_1 as a parameter in the equation 2.

I somehow have to tell SIMULINK to first solve equation 1 and after having solved that it should start solving equation 2.

[Timo Dietz]

If you insist on having a strict order and equation 2 must not run before equation 1 is done, you could e.g. do this:

Put equation 2 in an enabled subsystem and attach a comparator which checks for t_1 > 0 to the en input.

But, what's the issue with letting equation 2 run continuously and just rejecting all results before the output of your triggered sybsystem changed to t_1?

[Enea Baumann]

Hmmm, alright.

The thing is, equation 2 needs the whole output (so y(t)) of equation 1 and also the value t_1 from the begining of the calculation.

When I compute equation 1 and 2 simuntaniouly and continuously, the value t_1 is not available for the equation 2 from the beginning on.

In the picutres attached you can see the output of the triggered subsystem which is part of the equation 1. What I want is to used that value (0.65) as a constant in equation 2.

So I need to first calculate eqauation 1 which gives me the solution y(t) and also t_1.

Then I need to compute equation 2 with these to inputes (where y(t) is a function of time and t_1 is a constant).

Thank you so much for you help!!!

[Timo Dietz]

Does eq2 need the eq1 output starting at t=0 or at t=t_1?

[Timo Dietz]

The only solution for option 1 - I could imagine right away - is to duplicate eq1 with the argument t'=t-t_1 and put eq1' and eq2 (or eq2') inside the enabled subsystem. Thus, both would deliver the results as starting from zero and eq2 input t_1 woud be available, already.

[Enea Baumann]

Yes, eqation 2 needs the output from equation 1 startion from t=0.

Alright, yes that sounds like a solution. I will try that!

Thank you very much for your help!

[Enea Baumann]

I can't figure out how to change the argument to t' = t - t_1.

How would you do that?

[Timo Dietz]

I'm puzzled. You wrote eq1 is a y(t) function so I assumed one input is t, which could be easily changed to t-t_1. In case your eq is implicitly depending on t in terms of simulation time only, you don't have to change anything. Just put eq1 and eq2 in the enabled subsystem. Since enable is 1 after t_1 the internal equations start with the internal start time of zero, which is delayed by t_1 with respect to the overall simulation time.

Or are we talking at cross-purposes?

[Enea Baumann]

Sorry for that.

I was able to solve it. As you said I put eq1' and eq2 into a enabled subsystem.

The subsystem is enabled as soon as t_1 is reached (calculated by eq1).

The subsystem's input is t_1.

This way, inside the subsystem eq1' is computed feeding its output to eq2 and and t_1 is obtained by eq1.

Thanks for your help!

[Timo Dietz]

Perfect! This is what I tried to explain. Good luck, furthermore.