The Situation:
When trimming and linearizing a large Simulink model, we are consistantly running into memory problems with MATLAB that is causing it to crash.
Couple of notes:
It is a large simulink model with a lot of unit delay, delay, rate transition blocks
I have already went through the entire simulink sim and removed all data logging but I am still getting an error about "Data logging exceeded available memory" and "Cannot allocate sufficient memory for logging. Consider disabling logging or reducing the length of the simulation"
We are using two optimization algorithms with fmincon, 1. interior-point and 2. SQP, and trying multiple attempts to trim the model for one flight condition
We have a few sfunctions within the simulink model, two are large, the addition of one of the two large models to the simulink model that is being trimmed and linearized has made this issue worse
We are using fastRestartForLinearAnalysis
Reference the picture below for the error that I was able to catch
Now the questions:
Is there something with fastRestartForLinearAnalysis that is causing memory to be used up that we cannot see? (using the task manager on the pc, it doesnt appear that the full memory is being used/allocated)
Could this be caused by our choice of optimization algorithm? I have already read the "Choose the Algorithm" page which suggests that using SQP with this large scale optimization could cause issues
What else could it be?The Situation:
When trimming and linearizing a large Simulink model, we are consistantly running into memory problems with MATLAB that is causing it to crash.
Couple of notes:
It is a large simulink model with a lot of unit delay, delay, rate transition blocks
I have already went through the entire simulink sim and removed all data logging but I am still getting an error about "Data logging exceeded available memory" and "Cannot allocate sufficient memory for logging. Consider disabling logging or reducing the length of the simulation"
We are using two optimization algorithms with fmincon, 1. interior-point and 2. SQP, and trying multiple attempts to trim the model for one flight condition
We have a few sfunctions within the simulink model, two are large, the addition of one of the two large models to the simulink model that is being trimmed and linearized has made this issue worse
We are using fastRestartForLinearAnalysis
Reference the picture below for the error that I was able to catch
Now the questions:
Is there something with fastRestartForLinearAnalysis that is causing memory to be used up that we cannot see? (using the task manager on the pc, it doesnt appear that the full memory is being used/allocated)
Could this be caused by our choice of optimization algorithm? I have already read the "Choose the Algorithm" page which suggests that using SQP with this large scale optimization could cause issues
What else could it be? The Situation:
When trimming and linearizing a large Simulink model, we are consistantly running into memory problems with MATLAB that is causing it to crash.
Couple of notes:
It is a large simulink model with a lot of unit delay, delay, rate transition blocks
I have already went through the entire simulink sim and removed all data logging but I am still getting an error about "Data logging exceeded available memory" and "Cannot allocate sufficient memory for logging. Consider disabling logging or reducing the length of the simulation"
We are using two optimization algorithms with fmincon, 1. interior-point and 2. SQP, and trying multiple attempts to trim the model for one flight condition
We have a few sfunctions within the simulink model, two are large, the addition of one of the two large models to the simulink model that is being trimmed and linearized has made this issue worse
We are using fastRestartForLinearAnalysis
Reference the picture below for the error that I was able to catch
Now the questions:
Is there something with fastRestartForLinearAnalysis that is causing memory to be used up that we cannot see? (using the task manager on the pc, it doesnt appear that the full memory is being used/allocated)
Could this be caused by our choice of optimization algorithm? I have already read the "Choose the Algorithm" page which suggests that using SQP with this large scale optimization could cause issues
What else could it be? optimization, linearization MATLAB Answers — New Questions

​