Pentacopter build log

3689732611?profile=original

A step-by-step guide to building my own pentacopter:

https://people.csail.mit.edu/taodu/pentacopter_guide/guide.pdf

Everything (software and hardware) is open-sourced and publicly available.

flight video:

Firmware:

https://github.com/dut09/ardupilot/tree/holodeck-demo

Simulation software:

https://github.com/mit-gfx/multicopter_design

Frame design:

https://cad.onshape.com/documents/829cfef861a7d873d7f403de/w/2d807cb313793dabb7bfbee6/e/658cd3f46b668b90be89e8e8

E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones

Comments

  • I see. That makes sense. Thank you for your clarification!

  • Cool. Thanks. As you know, EKF is a LQ estimator, so adding a LQ regulator = LQ controller = stochastic controller.
    I mistakenly assumed you’d worked one into ArduPilot, but it’s your sim. Still, nice to see!
  • Hi Auturgy,

    Not sure I understand what a 'stochastic controller' is. I guess you mean the LQR controller in our old paper demo and in the software. The code of our LQR controller is here:

    https://github.com/mit-gfx/multicopter_design/blob/master/projects/...

    I wish I could have a writeup for it too if I had enough time.

    mit-gfx/multicopter_design
    Contribute to mit-gfx/multicopter_design development by creating an account on GitHub.
  • Hey John Hestness,

    I haven't tried it that sounds like a really interesting idea. I tend to believe it can survive the loss of one motor. Of course, the controller code needs some modification accordingly. In ArduCopter, I guess it could be something like this:

    - Added a new flight mode in ArduCopter.

    - When the new flight mode is triggered, turn off one motor and update the roll, pitch, yaw, and throttle arrays in AP_MotorsMatrix.

    You are right, the five rotors are flat and level, and it was indeed a bit inefficient. This was reflected in Figure 10 in our paper (https://people.csail.mit.edu/taodu/copter/copter.pdf). The 'optimized' version has five tilted rotors.

    https://people.csail.mit.edu/taodu/copter/copter.pdf
  • The pentacopter is interesting, but you’ve implemented a stochastic controller too? That I’d really like to see a write up on!
  • Cool project, great write up, and thanks for sharing! Can a pentacopter survive the loss of one motor? Have you tried it? Do you have another reason for using five?

    I see you mentioned that you did calculations for using tilted rotors, but the rotors in the picture appear flat and level with the frame, is that correct? If so, then the two ccw rotors would have to spin faster than the three cw rotors to cancel out torque? I suppose that would introduce a small inefficiency unless you sized the motors differently? Thanks. 

This reply was deleted.