Shrediquette - the award-winning multirotors by William Thielicke

I got a lot of requests for the matlab source code of my tracking algorithm. (If you wonder what I am talking about, watch this video of the tracking process via Matlab: http://vimeo.com/6347100 ) As you will see, it is not that complicated. I did not upload the complete source code, because that doesn't make too much sense. My GUI for example, contains a lot of code and it is most likely irrelevant to most of you. I left out the calculations of the different parameters for my PID control loop too, they will make things appear more complicated than they really are. Here comes the code: Matlab Tricopter Tracking If you have questions, I would suggest that you put them into the comments for this post. Doing so, more people might benefit from it.

...a new movie showing some of the capabilities of my copter design. This is supposed to be a somewhat less technical movie... Enjoy! Tricopter - The Movie... from W. Thielicke on Vimeo .

Next Steps...: I will also post the circuit layout, the source code of the copter and everything else on this webpage soon.

Finally, it's done... My tricopter flies along a three-dimensional path without me doing anything. The waypoints are recorded by some sort of "teach in" (basically taking snapshots of different positions in space, the flightpath is hence calculated as the cubic interpolation between these snapshots).

A mobile phone (K800i) controls the tricopter. The phone is linked via bluetooth to the laptop.

This movie demonstrates how the tricopter hovers autonomously inside a room. There are three LEDs fixed to the copter which are filmed by a webcam from below. A Matlab script then makes the copter hover in place.

Improved algorithms and a more confident operator...

Matlab displays the attitude of the copter live and in 3D (and in colour ;-D !!)

Due to the improvements mentioned above, I was able to do the very first looping...

Some changes...: I got new motors (Scorpion s2212-26) and new gyro's (ADXRS610). The ESC's were modified for fully digital I2C control. Additionally, I rewrote most of the control loop (finally, I understood what I actually have to control to make stable flights possible). This movie demonstrates two different control modes (control of the angle vs. control of angular velocity).

... interestingly, decreasing the size of the tricopter did not have any noticeable effect on flight stability.

Smaller = easier to transport...

Flying around in the parc... The control loops are still not optimal.

A flight through my balcony door.

This is, by the way, how my board looks like from below. Beautiful chaos... but still functional. I am planning to make a PCB soon...

... and this is how it looks like in flight. Still not 100% stable, but I seemed to be on the right track.

... electronics installed...

... the new frame...

A 3D model of prototype 2...

Some scenes of the first prototype.

First release of my "tricopter debug tool". It displays all the parameters that are relevant for a (potentially) successful flight...

Birth of the tricopter (03.11.2008)...: The second flight of the first prototype... Not very stable yet. But at least it seemed to work after all!