Why there is a time delay of one time step while co-simulating a FMU in Simulink?

Question

MathWorks Support Team 2021-1-19

1
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/720854-why-there-is-a-time-delay-of-one-time-step-while-co-simulating-a-fmu-in-simulink

编辑： MathWorks Support Team 2022-6-16

I exported an FMU from a Simulink model with Simulink Compiler and imported it into Simulink with FMU Import block.

In the results, when comparing the Simulink output and the FMU output, I see that they do not match with a one-time step difference/one-time step delay.

What is causing this behavior?

请先登录，再回答此问题。

Answer 1

MathWorks Support Team 2022-6-9

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/720854-why-there-is-a-time-delay-of-one-time-step-while-co-simulating-a-fmu-in-simulink#answer_601054

编辑：MathWorks Support Team 2022-6-16

The delay you see in co-simulation is a documented behavior for co-simulation of FMUs and standardized in the FMI standard.

Here is a reference in our documentation:

https://www.mathworks.com/help/slcompiler/gs/export-simulink-models-to-functional-mock-up-units.html

"You might experience an expected time delay in the exported FMU for Co-Simulation mode."

Due to co-simulation, there must be a time delay for communication between the local solver of the FMU and the Simulink solver. This is analogously the case in other simulation software when the FMU is co-simulated.

Here you can find more information about this and also ways how this can be intercepted under certain circumstances with special FMUs (which have certain properties) by so-called Numerical Compensation:

https://www.mathworks.com/help/simulink/slref/using-numerical-compensation-for-co-simulation-integration.html

By the way, you define the step size of this delay via the Communication Step Size: