At the moment I am totally dependent on PID Algorithm to stabilize my Quadrotor. It seems slow to react and to solve errors. If I increase the P co-efficient of PID it becomes highly unstable. I am using Arduino MEGA as my main controller to look after the Controlling Algorithm. Can anyone suggest a more effective control system which can run on my ATMEGA 2560?
Dear Bhaumik
Hi, It might be helpful having a short look at recent published paper in IEEE Control Systems magazine, Apr 2013 Volume 33 Number 2, with title "An Educational Setup for Nonlinear Control Systems". You may be able to review it through this address, www.ieeecss.org/pab/csm. It discusses on Quadcopter as its nonlinear system's example and how to control it.
Kind regards.
I suggest you using Optimal Control or Robust Control to increase reaction pace and deplete steady state error; Also you can use a combination of these controllers with you PID
I think the best controller is sliding mode controller (SMC) for this purpose. since it has fast response, zero steady state error, and no over-shoot. the best point of this type of controller you do not need to know the exact Transfer Function of your system. Finally, you can use it with ATMEGA 2560 micro- controller easily.
I agree with Mr. Hazim Tahir. SMC has a fast response with zero steady state error. For your purpose, it is the best.
Hi Baumik, you must firstly establish the goals for the closed-loop control system of the quadrotor. There are a lot of control schemes using backstepping, feedback linearization, sliding modes control, fuzzy control and Takagi-Sugeno fuzzy control, just to mention a few. Really, most of the time we must sacrifice some aspects for the performance in order to design the control system. More even if the task will take place on an indoor or outdoor environment. Or maybe if you want that your quadrotor be oriented to aggresive/acrobatic maneuvers or low energy consumption. A lot of control schemes relying on the combination of some of the control techniques above mentioned are reported in the literature.
A thesis entitled "TRAJECTORY GENERATION AND CONTROL FOR QUADROTORS" by Daniel Mellinger, Phd 2012 from University of Pennsylvania, you can search the web and find it, i think it helps you.
P+D usually works well unless you need a highly maneuvrable craft, then you need one of the more advanced methods. Have you separated out the attitude loops from the position? You should not need any integral action. If you increase your P gain you must also increase your rate feedback to increase the damping and so maintain your stability margins. You can increase it until the rotor demands start saturating the rotors (max speed or stall).
I agree with Mr. James Whidborne. P+D works well, and integral action may saturate the rotors.
The Adaptive Fault Tolerant method can be the best method, Because Sliding Mode is so robust about external error but not about intrnal problems. This algorithm helps you to handle both internal and external errors.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Control Theory
- Control Systems