Navigation development

[nate-roy]

• [ ] Performance simulation
• [x] Landing point selection/obstacle avoidance validation
• [x] Software-limited turn rate validation
• [ ] Locomotion validation

  • [x] Euler verification

• [ ] Wind correction validation

  • [x] Software validation
  • [ ] Flight test validation

• [ ] Landing point orbit validation

  • [ ] Software validation
  • [ ] Flight test validation

[mallamacimj]

9/11/24

Did this week

• Nothing lol

Next Week

• Euler Validation
• Look over the wind correction code from last year
• Move TADPOL code over to new MMFS

9/18/24

Did this week

• Wrote MMFS orientation propagator: https://github.com/Terrapin-Rocket-Team/Multi-Mission-Flight-Software/tree/imu-relative-atittude/src/Sensors/IMU
• Orientation Visualizer: https://github.com/Terrapin-Rocket-Team/Payload/blob/orientation-visualization/Simulations%20%26%20Team%20Tools/orientationVisualization.py

Next Week

• Use the orientation visualizer to validate euler angles - think a 321 should work for yaw???
• Look over the wind correction code from last year
• Move TADPOL code over to new MMFS
  • Review and merge in unit testing PR: https://github.com/Terrapin-Rocket-Team/Multi-Mission-Flight-Software/pull/51

9/25/24

Did this week

• Rewrote MMFS orientation propagator to be better: https://github.com/Terrapin-Rocket-Team/Multi-Mission-Flight-Software/pull/62
• MMFS 2.0.0 release (this is a stable version to work on)
• MULE parts at terrapin works for water jetting... prob be done around Monday

Next week

• Move TADPOL code over to new MMFS (2.0.0)
• @nate-roy give me another thing to do

[ChenJeff123]

9/11:

Didn't get anything done. Next week will look over old descent path code and integrate wind correction code

9/18:

Brushed up on descent path code. Looked over wind correction code - need some questions answered. Will hopefully fully integrate next week.

9/25:

Integrated wind correction code into descent simulator. (Edit: doesn't work as intended still). Need to add graphs with change in velocity over time, as well as functionality to plot multiple graphs at once with different variables.

10/2:

Finally has functional wind simulation. (Not fully tested, could have bugs still) Now have graphs of vehicle velocity over time in x and y axes. Need to test further and add functionality for multiple overlayed plots with different parameters.

10/9:

Overlayed wind corrected and non wind corrected graphs. (They aren't exactly one-to-one due to different randomly generated wind). Also, the velocity graphs are for the wind corrected version.

10/23:

Most of descent simulator is finished: ability to add run simulations of multiple starting points, with/without wind correction. Working plots of position, velocities, and heading. Still sometimes messes up though. Still need to do: incorporate wind estimation.

10/30:

Wind estimation is still in progress :((. Hope to get that done by next week.

11/6:

Still working through some issues with wind estimation.

11/13:

Wind estimation have been incorporated.

[rowleyhc]

10/28: Finished writing and testing some basic navigation simulation that incorporates the angle limitation command: From playing around with different targets and starting headings/points, it seems like the goalCommand function(which effectively limits the difference between the goal and actual angle that the controller uses) is working properly, and doesn't cause any issues. However, as can be seen the the heading is still changing very rapidly(about 180 degrees in 0.1 seconds, which seems to be the maximum the controller will output(1800 degrees/second)), which is still going to cause some issues. This seems to be a limitation of the controller function, rather than the goalCommand functionon. A rough band-aid solution for now seems to be just multiplying output pwm by 1/20, which produces a much more reasonable turn rate: To do: figure out a better way to limit the turn rate so that the pwm output still allows the motor to be more responsive and isn't unnecessarily scaled down

10/30: Rewrote goalCommand function so that turn rate is fully limited by it and pwm can be left unchanged:

11/13: Started working with Simulink, made a simulation with the same capabilities as the python one I made earlier, though using the functions for angular acceleration Michael generated from the rotation testing. Moving around in a rough circle(some variation suggests further PD tuning needed):

To do: -Keep troubleshooting edge cases that seem to break the model(a lot of this just seems like error due to how Simulink solves the system, which seems very sensitive to the tuning of the PD controller and how close the vehicle is to the target) -Incorporate more realistic motion simulation by modeling the parachute connection as non-rigid and with forces -Eventually, model wind, which would affect the vehicle's motion and also move the center of the circle(the rocket)