How Do You Convert Drone Position Errors into Roll and Pitch Angles?

AI Thread Summary
The discussion centers on developing a flight controller for a drone, specifically how to convert position errors in the x and y axes, along with yaw, into desired roll and pitch angles. The participants express confusion about the necessity of roll and pitch angles when the movement is primarily horizontal. They clarify that they are implementing a cascaded control setup, where the outer loop uses x and y errors and measured yaw to compute the desired roll and pitch angles. A request is made for guidance on the formulas needed for these calculations, as well as for a block diagram of the control scheme. The conversation emphasizes the importance of understanding the overall control strategy before delving into specific equations.
Young_kids
Messages
3
Reaction score
0
TL;DR Summary
computing remaining 3D angles from x, y and the angle in between them
Hi guys
We have a problem we we are building a flight controller for a drone, but fail to understand how to construct the position controller.
It takes the x, y positions and the angle in between them which we call yaw. How to compute the angles roll and pitch for the desired x and y positions. We have an idea of the following
1580113421740.png

Can someone explain what formula that converts the x error and y error as well as the angle yaw to the desired roll and pitch angles.

We found a formula

1580113599756.png

and
1580113623166.png

But we don't understand why, hope you can help
 

Attachments

  • 1580113087821.png
    1580113087821.png
    2.2 KB · Views: 241
Physics news on Phys.org
Young_kids said:
It takes the x, y positions and the angle in between them which we call yaw. How to compute the angles roll and pitch for the desired x and y positions.
If the direction is just horizontal (x & y), why do you need the other angles?
 
A.T. said:
If the direction is just horizontal (x & y), why do you need the other angles?

Because we are making a cascaded control setup, the inner loop controls, pitch, roll, yaw, z. As this is an underactuated system, the desired pitch and roll angles are obtained by cascaded controllers, which make up the outer loop. This outer loop takes the x_error, y_error and measured yaw, and should compute the desired pitch and roll angles in order to achieve the desired position. We are unsure how to compute these desired desired pitch and roll angles, can you give us a hint please
 
Young_kids said:
Because we are making a cascaded control setup,
Before looking at specific equations, the overall plan for cascaded loops must make sense. Can you post a block diagram of the entire control scheme from sensors to actuators?
 
anorlunda said:
Before looking at specific equations, the overall plan for cascaded loops must make sense. Can you post a block diagram of the entire control scheme from sensors to actuators?

Hi anorlunda
Thanks for replying. Our block diagram is as follows overall
1580131382071.png

We are unsure of how do the world to body conversion, where the x_error, y_error and yaw are used to compute the desired roll and pitch angle
 
Due to the constant never ending supply of "cool stuff" happening in Aerospace these days I'm creating this thread to consolidate posts every time something new comes along. Please feel free to add random information if its relevant. So to start things off here is the SpaceX Dragon launch coming up shortly, I'll be following up afterwards to see how it all goes. :smile: https://blogs.nasa.gov/spacex/
Thread 'SpaceX Starship development: 7th flight January 10'
Watch the progress live This is a fully stacked Starship (top) and Super Heavy (bottom). A couple of too-small-to-see cars near the bottom for scale, I also added a Saturn V and the Statue of Liberty for comparison. 120 meters tall, about 5000 tonnes when fully fueled. Twice the mass and over twice the thrust of Saturn V. The largest rocket ever built by mass, thrust, height, and payload capacity. N1 had the largest diameter.[/size] But its size is not the revolutionary part. It is designed...
Venus does not have a magnetosphere, so the Galactic Cosmic Rays (GCRs) environment shall be much worse than in a LEO environment. Looking to the std radiation models for Venus, the standard radiation-hard space level electronic component with tested immunity LET = 85 MeV-cm2/mg seems not enough, so, for example, a 1cm2 Si die will suffer considerable flux above this level during a long mission (10 years for example). So, the question is, usually we are not paying attention to latch-up...
Back
Top