Speed up Simulink model which uses a Matlab function block

Hello,

I would like to speed up the running of my Simulink model that uses a Matlab function block. Without that function block my model runs fine and fast. When running the function in isolation, I timed it to 0.07s using tic/toc. But when I include the function block in my Simulink model, it runs very slow (even when the function block is not connected to anything).

Is that a normal behavior or is there anything to improve?

Thanks!

Answer 1

[Dhruv]

Hi @Matt

I understand that upon including MATLAB Function block in your Simulink model slows down the execution time. This behaviour could be because of the function that is implemented in the script which might be taking more time because it is not optimized. Here are some strategies to improve the performance:

1. Vectorization: Ensure that your MATLAB function is vectorized. Avoid loops where possible by using matrix operations, which are generally faster in MATLAB.
2. Code Optimization: Use the MATLAB Profiler to identify bottlenecks in your code. Simplify complex operations and preallocate memory for variables to avoid dynamic resizing.
3. Data Types: Use fixed-point data types instead of floating-point where possible. This can significantly reduce computation time, especially for embedded systems.
4. Sample Time: Check the sample time of the MATLAB Function block. A smaller sample time can lead to excessive computations. Adjust it to match the needs of your model.
5. Parallel Computing: If you have a multi-core processor, enable parallel computing features in MATLAB, which can distribute computations across multiple cores.

By applying these strategies, you should be able to improve the performance of your Simulink model.