Step input to 6DOF impacts performance

3 views (last 30 days)
Thomas
Thomas on 8 May 2024
Answered: Pooja Kumari on 8 May 2024
I am currently using a 6DOF block with the moment input being a sum of two signals. When one of these signals is a step input (i.e., going from zero to a nonzero value over a single timestep), the simulation slows down significantly; however, this only occurs for a moment about one axis and not about the other two axes. Are there any potential solutions for this?

Answers (1)

Pooja Kumari
Pooja Kumari on 8 May 2024
Hi Thomas,
When you encounter a significant slowdown in your simulation due to a step input in one of the moment axes using a 6DOF block, it suggests that the simulation is having difficulty resolving the sudden change in conditions.
  1. Review and Adjust the Algebraic Constraint Method Settings:Given that the 6DOF (Quaternion) block relies on the algebraic constraint method, which is sensitive to the gain value settings across time step.
  2. Adjust Gain and Time Step: Fine-tune the gain settings and reduce the time step size around the step input to improve handling of sudden changes.
  3. Smooth the Input: Gradually transition the step input using a ramp or sigmoid function to reduce the simulation's computational strain.
  4. Optimize Solver Settings: Experiment with different solvers and adjust tolerances. A variable-step solver may better manage the abrupt input.
  5. Simplify the Model: Ensure the model's complexity is necessary for the simulation goals. Overcomplexity can slow down the response to step inputs.
You can refer to the below documentation for more information on 6DOF:

Categories

Find more on Manual Performance Optimization in Help Center and File Exchange

Tags

Products


Release

R2024a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!