Practical application and background of PID algorithm for UAV

Intuitively, this is an enlarged version of the four-axis UAV. In my opinion, the technical difficulties faced by the technical team of Yihang are the combination of software and hardware and automatic control, which is where they spend a lot of effort. If the quadcopter only discusses the aerodynamic principle and structure, it is not even as complicated as the fixed wing and the general helicopter. It can fly freely in the sky by relying on the lift generated by the motor group and propeller to offset and adjust various attitude forces.

Although this process is easy to describe, in fact, the quadcopter is not so much dancing with the air as fighting with it. The beauty of the unmanned aerial vehicle lies not in its elegant curve in the sky, but in its lack of complex mechanical structure, total distance hub, control surface and server for controlling flight, and its structure is simplified to only the driving motor connecting rod and control circuit, which is exactly like a personal aircraft and a flexible vehicle in the future, but it belongs to a new vehicle that is still being recognized and developed by the public.

At first, if you want to fly according to this structure, the pioneers of aviation will laugh at you and go to Mars. An aircraft like this is simply unreasonable. I wonder how many people have done people who want to ascend to heaven with engines and rockets attached, and most of them have paid the price of their lives. However, with the gradual summary and development of automatic control theory, outstanding scientists and engineers can use computers to realize the output adjustment of millisecond-microsecond level through mathematical ideas (limit approximation, infinite division-calculus) in a very short time. For example, when these four motor control inputs are output, it usually takes 300 parameters (there is a graph in this paper to describe how PID algorithm corrects the output value, So that it can respond to external interference and tend to the target value, so that it can fluctuate up and down around it. A good flight control system can know by looking at the curve.) So many parameters are so complicated that people mistakenly think that they can't be realized for a long time, but the actual situation is the opposite, and the solution is engineering.