Hello. I aim to model a battery assist for a hybrid SUV. The assist harvests power from a TEG. This power is conditioned with a two-stage boost converter, which is also subject to a P&O algorithm. The converter's output is fed into the SUV's hybrid battery, attached to which is a current drain representing the car's electrical system plus inefficiencies. When I run the model, I find that three hours are required for the model to go from 0 to 99% completion, and then an additional two hours are required for the model to terminate. Upon termination, I am told that there is a memory allocation error, and I am not able to obtain any scope data. Based on debugging, I am reasonably certain that the main memory sink is in the two-stage converter, as running the model without this element or the MPPT speeds the entire runtime process up to one hour. I have asked the MathWorks help desk for advice; however, they were unable to assist me. It would be great for me if someone would be able to recommend methods for optimizing memory usage in the boost converter block of the simulation?
Your help is much appreciated. Thank You in Advance.
hi,
use Matlab function to write the equation for your converter, it is better not to use Simulink blocks. designing converter through mathematical equations will take less memory and faster sampling.
you can also try to run the model after changing the solver and sampling time.
Dear colleague,
I agree with Jitendra Sing, additionally setting the "power GUI" to "Discrete mode" and using variable step solver configuration, will significantly increase simulation speed.
- Badges
- Science topic