Use WLTC drive cycle data instead of step input

I’m currently working with the PMSM thermal model. I replaced the input speed (RPM) with custom WLTC speed data, specifically the first 50 points of the WLTC profile. I adjusted the simulation time to 50 seconds accordingly. I also interpolated my data to have a time interval of 0.01 seconds, and I verified the input RPM values using a plot—everything appears normal.

However, I encountered the following issues:

1. Torque Not Measured: In the simulation results , the demanded and achieved torques are not being measured.
2. Temperature Explosion: After around 10 seconds into the simulation, the motor temperature exhibits an unexpected and unrealistic spike or explosion.

I’ve carefully checked the input RPM data, and everything seems fine. What could be causing these issues? How can I resolve the torque measurement issue and fix the temperature explosion problem?

Any help or suggestions would be greatly appreciated!

[Shivam Gothi]

Hello @轩昂,

Can you please provide the "custom WLTC speed data" ? It will be helpful in reproducing the issue and suggesting a proper solution.

[Sabin]

Can you share the model, not easy to investigate without a model? At first look it seems that the system is unstable but not exactly clear what may be the root cause.

Answer 1

[Sabin]

What may be happening is that the library example is set up for operation below the base speed (Zero d-axis control). If the RPM profile that is used as reference goes above the base speed, it is quite likely that the voltage limit is reached and is saturated for quite some time, and the controller will eventually become unstable. Also using individual switching devices in the power converter with long simulations (50 seconds) will probably cause slow simulation. May be an option considering average value inverter or even DQ-only simulation such as this example:

https://mathworks.com/help/sps/ug/simulate-pmsm-drive-with-thermal-model-dq-frame.html