Incredible Heavy Lift Quadcopter (HLQ) Engine PID Tuning

This video gives a glimpse into the process we went through to implement and tune the PID controllers that keep the engines at constant speed throughout varying pitch angles. The video was taken in a early phase of the process, and at extreme conditions where the step sizes were very large. Because of this, you will notice some oscillations which have since either been tuned out, or are at an RPM level that we will not fly in, and therefore are not a problem.
You may also notice some delay in the reaction of the controller on the larger step inputs to the blade pitches. These types of large and sudden changes in pitch are not expected during flight (at least not in all 4 rotor heads at once), and will actually be avoided (slow and steady wins the race when it comes to HLQ). The delay in reaction was minimized through tuning, but because these conditions are extreme, their existence was deemed acceptable at these extreme conditions. We have later video of flight controls testing (coming soon to KZbin) showing that in flight like conditions, the engine rpm is very stable, and these delays were undetectable, at least by human ears.

[@samualwhittemore228]

Keep up the good work....Awesome test!

[@Jped277]

When do we get to see it fly? I've been waiting for months lol. Looking awesome tho.

[@sudo9445]

me too!

[@T70781]

Your videos would be much more interesting if you included commentary regarding what you are learning and how each phase fits into the overall plan. As it is, I'm losing interest until you actually fly it, if that happens.

[@Throzencrew]

Take a look at the Speeduino project. It's arduino based engine management and it can be used to control the fuel, also you can use it's closed loop idle control to control a constant RPM, instead of using servos. speeduino.com

[@ericperkins3078]

I think you have overlooked a very basic problem when it comes to scaling technology. The reason that small electrically operated drones are so successful is because of their small size and the relatively light weight of their components. Because the small, lightweight electric motors and propellers can respond quickly to the nuance of electronic input, they can be controlled very rapidly and precisely. As you scale their size upward, the sheer weight and thus inertia of the components reduces the ability of the electronic controllers to make changes quickly enough to make control practical. When you get to the size of copter you're working on and have moved to ICE power, you have lost nearly all of the nuance of control and all the advantages of light weight. The ICEs cannot and will not respond with the speed and accuracy of small, powerful electric motors. The additional heat (both combustion and friction) and vibration are stupendous compared to smaller, electronically controlled electric motors. I salute your efforts nonetheless. You are literally having to "reinvent the wheel". Once you have perfected the design (assuming you can, and I hope you do, properly patenting everything along the way!) you should end up quite famous and wealthy, as well.

[@Incrediblehlq]

The inertial energy of the system is overcome primarily by utilizing variable pitch rotor blades. In this way, we can maintain a constant RPM, and our control reactivity is limited only by how quickly the servos can actuate across their throw. As it turns out, HLQ has quite a bit of control authority, as can be seen in the video we just uploaded here: kzbin.info/www/bejne/mZSzd5aDo9isn7s See the second part of the test called where we aggressively (comparatively) test the flight controls.

[@RahulPatel-qz2vb]

Fuking this thing old 5 years but still no any update of flying

[@allieddesigncollc8377]

Hi, what engine are you using?

[@randysonnicksen9475]

Are you planning to use any "feed-forward" in your engine speed loop? I am working on a similar project but not as far along. My thinking is that I will use a "standard" flight controller (for electric quads), but have an additional speed controller like you, but I will take the 4 rotor pitch signals and sum them together, and use this as a a feed-forward signal to the engine, so that power is added as pitch is increased. If the feed-forward gain is properly selected, this will greatly reduce the magnitude of speed disturbances. I plan to characterize the pitch/torque relationship before summing the 4 pitch signals, and also characterize the throttle/torque curve at flight RPM to attempt to eliminate RPM deviations caused by rotor pitch fluctuations. From your video it sounded like you are just using straight PID. I could hear the motor slow down as load was added, the ramp up, overshoot, and then settle in back at sepoint. Need to be VERY careful with low RPMs for several reasons. 1)centrifugal force of the blades (which is what holds a typical helo up) is proportional to RPM^2. a 10% loss in speed is a 19% loss in centrifugal force. 2) similar rule applies to lift. Lift is proportional to velocity^2. So if the RPM drops as pitch is increased (because you are wanting MORE lift), you might actually get LESS lift if the motor slows down too much. Then the flight controller would ask for even MORE pitch......You get the idea. Not long before the quad and ground would be introduced to each other. The point is - You need really tight control on RPM. Probably better than what you displayed on this video. Good luck. Let me know if you want more info on how to implement "Feed-Forward".

[@davidgarcia-mendia2137]

Hi there, which ESC do you use to control the throttle of the gasoline engine?

[@Incrediblehlq]

There is no ESC needed. It is controlled using and PID running on an arduino and a servo to manipulate the throttle valve.

[@heliscat]

Electronic swash plate. Same exact concept. I think this is why gasser helicopters were so popular in the heavy lifting department years and years ago; their sheer lifting ability. Looks like a fun little project guys, great work.

[@bikemike2777]

Think your running a Lil lean on fuel and also I could recommend amsoil Saber mixed at 50.1 and from a previous vid you had with the destroyed piston. The top of you piston had indicated that it looks like detonation was occurring may want to keep tabs on the fuel you run in small engines. Love your project choppy

[@nguyentrongminh3058]

Name engine?

[@CrArC]

I am dying to see this thing in the air!

[@knarFkcalB]

At least one of the motors was going into resonance vibration as RPM changed. That can't help performance or lifetime. The mounting needs to be modified to dampen that out.

[@0nec00ler]

Add more derivative! Lol :)

[@aGRUMPYCAT]

nice, progress!

[@Potato22]

KEEP ON!

[@CullenCraft]

Awesome work! As fuel is used, wont the HLQ lose balance?

[@ultrastoat3298]

Physically certainly. But with 4 independent rotors. It's easily compensated for.

[@CullenCraft]

UltraStoat But they aren't independent. Thy are all sharing engines. I suppose the pitch of the blades could be adjusted to compensate but that will effect maneuverability.

[@ultrastoat3298]

DedElec Well, yes, their RPMs are linked and not independent. But their thrust is. And really, that's what matters. I doubt any affect on maneuverability can even be measured considered the small amount of mass involved. Even if the amount of mass was larger, your maneuverability should be pretty good unless that imbalance would cause certain maneuvers to reach maximum thrust (ergo max throw on the servo), when otherwise it would not have. All of this will be blended out by the flight controller like it wasn't even there.

[@gilbertsedo]

maybe a circular shape gas tank with a concave bottom placed in the middle will solve the balance issue. maybe?

[@sergioorjuela7896]

looking good