Use spreadsheet time input for simulation timestep?

Hi all,

I am importing data from several spreadsheets, as would like to use the Time column of one of them to act as a 'master' time from which the simulation will be run from. In other words I'd like the time output of the simulation to exactly match the time column of the spreadsheet.

How would I go about implementing this?

Thanks

[Adam Danz]

What do you need help with (you haven't asked a question)?

Here are some tips on how to get help here:

https://www.mathworks.com/matlabcentral/answers/6200-tutorial-how-to-ask-a-question-on-answers-and-get-a-fast-answer

[Gazebo]

I need help with using a spreadsheet time column as the simulation time.

My question would be

How do I use a spreadsheet time column for the simulation time?

Many thanks

Answer 1

[Fangjun Jiang]

Theoritically you can't specify the exact simulation time as it is determined by other factors such as accuracy or tolerance. The time vector in your Excel file could have non-zero start time, large time step, etc.

Practically, you can set up your solver to match your desired time vector, if it is starting from zero, having a fixed time step.

More generically, you can run your simulation, get the output, and then re-sample the output to match your time vector. See Simulation, Model Configuration Parameters, Solver and Data Import/Export.

[Gazebo]

Thanks for the help. The main problem with your first suggestion is that the time vector in the .csv is not fixed step (ranges between 0.2 to 0.7s). The other issue is that the spreadsheet also contains triggers ( one column is a boolean flag that changes the code execution, so sampling this value twice or skipping over it would result in an unaccurate simulation, even if I resample the output).

I will continue to explore options, thanks.

[Fangjun Jiang]

I would treat the column of boolean flag as one of the inputs. Use this input to trig the execution of some Simulink subsystems. So this has nothing to do with time vector or sample time.

[Gazebo]

Apologies, I meant that the main issue is the result of the previous execution is fed into the next ( I need to use a lot of memory blocks).

[Fangjun Jiang]

You can save the output and state of a simulation and use that to continue another simulation. Or, if the simulation is confinuously running, use Merge block to merge multiple conditionally executed subsystmes. Anyway, trying to specify the simulation time vector seems in the wrong direction.