You could use two Matlab-Function blocks, where one returns the actual cpu time on every call, while the second one returns the cpu time on the first call only. The difference of these two values gives you the real elapsed time in seconds. Comparing the elapsed time to a constant (stall time limit) can be used together with the stop block to cancel the simulation before the integration has finished.

Hope this helps to solve your problem.

Best, Rouven

Here is a solution essentially similar to Rouven Christen's, but with only one Matlab Function block.

If your optimization is iterative you can limit the process by the number of allowed iterations through an iteration counter. You can also set your allowed error to higher values such that if this value is reached the optimization stops.

Best wishes

ok thanks for you, thanks for guidens

Perhaps sth like Simulink® Real Time Execution could solve your problem :

https://www.mathworks.com/matlabcentral/fileexchange/21908-simulink®-real-time-execution