Missed ticks meaning in Simulink Desktop Real-Time

Hi,

I have a really simple question. In all the Simulink Desktop Real-Time library I can find the word "tick", but I can't find an official explanation of its meaning.

I assume that a tick is the time length of the discrete solver, but for my real-time application I just want to be sure it is, because I need to evaluate the lag in millisecond the software could reach. So if it's correct I would just multiply the number of missed ticks lagging for the step length.

Thanks in advance.

Answer 1

[Jan Houska]

Hi Federico,

yes, your understanding is almost correct. A "tick" is an event when the real-time kernel tells a block that the time to execute is now. That is, if you miss one tick, the block is one sample period behind. But please note that this is related to the sample time of the particular block that provides the Missed Ticks output, not to the per-model discrete solver time step length. Although these are the same in many cases, this is by no means a rule and they can be and often are different.

Good Luck, Jan

[Federico Veronese]

Thank you Jan, so you are telling me that if the block says that is lagging by 1 tick, it means that it's for 1 of ITS sample time behind, and not for a sample time of the model (discrete solver) behind.

If my statement is correct I could find the time the system is lagging just by multiplicating the number of missed ticks of a block for the sample time of that particular block (and not for the sample time of the whole model/solver).

Thanks in advance, Federico.

[Jan Houska]

Yes, exactly.

Good Luck, Jan

[Khalid Hersafril]

I have a general question just to understand the general idea behind ticks. So I am currently using Normal mode for Simulink Desktop Real time, and as seen in the figure above, there is some missed ticks. For this scenario, as far as I understand, it falls under the following condition:

Process Contention — The model generally executes faster than required to keep up with the kernel. Process contention or some random operating system condition prevents Simulink from executing the model over some time period. In this case, the number of missed ticks jumps to some number, then missed ticks decreases to zero as Simulink catches up with the kernel. You can identify this situation by a sawtooth-like shape on a Scope connected to the Missed Ticks port. (Source: Simulink Real time Connected Mode)

So, since my model executes faster, it "stops" for a moment which led to a huge step occuring here. It then gradually decreases to zero which means that the simulink catches up with the kernel.

In terms of calculation, since the missed tick here is 2, and if my time-step is 0.005s, does that mean my block is behind the kernel by 2*0.005s=0.01s for that particular time-step?

Is it also safe to assume that the model is also behind the kernel by 0.01s as well since the block would still be executed in the same time loop as the model?

[Jan Houska]

Hi Khalid,

yes, your computations are correct. 2 missed ticks mean that you are 0.01s behind the kernel if your sampling period is 0.005s.

As to the block vs. model question, the answer is generally yes - if the block is 0.01s behind the model, we can generlly say that the whole model is.

This does not account for the situation where part of the model has been already executed before Simulink has been temporarily suspended by the OS, so actually portions of the model may not be that much behind. However this situation would typically only affect internal states of the model anyway and would not visible to external observer, so I think it can be safely ignored.

Good Luck, Jan