Remus Pitching Down - Thrust Step Input

[BradenMeyers]

I was working with the the REMUS 100 simulation tool. I put a step input on the thruster to full thrust and observed a pitch downwards.
After some debugging it seems to be due to the dissipative forces being calculated.

I have a branch that shows the few lines I modified to highlight this unexpected behavior.

https://github.com/BradenMeyers/PythonVehicleSimulator_DrewUV/tree/pitch_remus

If this is normal behavior is there an explanation for why this would happen?

[cybergalactic]

First of all, thank you for adding the torpedo vehicle to the simulator. I have merged the code, and if you make further updates, please feel free to submit new merge requests. Please also add yourself as an author in the files.

Regarding the non-zero pitch angle for a step input in propeller RPM, I first tested the SIMremus100.m (calling remus100.m). In MATLAB, there was no pitch bias. Then, I identified the difference: applying a step input of 1525 RPM can result in numerical stability issues. It is unphysical to change the RPM instantaneously. In the MATLAB code, and also in remus100.py, there is a time constant self.T_n = 0.1 seconds (which I have updated to 1 second). This acts as a low-pass filter for the command, helping to avoid numerical issues caused by the large step input.

Additionally, in MATLAB, I use the RK4 method to integrate the dynamics. I recommend using RK4 instead of Euler’s method in the Python code as well, though this will require an update.