Miscellaneous PID control algorithms-Article 2: tuning & #183; Three Quantities

In the previous article, the three constants in the PID algorithm play an approximate role in the PID algorithm, but in actual use, how to adjust the three PID control algorithms (or set them in more professional words. First, you can set all the three constants KP, KI, and KD to one and check whether the system's adjustment is too fast. Roughly determine the number of places that need to be shifted to the right of the output control result for the final control. Then, we set KI and KD to 0, and KP will gradually increase the test from 0 until the controlled amount has been out of tune, and there will be a small fluctuation. At this time, the kp is almost adjusted. We can continue to adjust the Ki, and increase the Ki so that the value of the controlled variable can finally be set as much as possible. The adjustment of the Credit quantity ki is similar to that of the Kp, from small to large. However, note that the increase of ki will increase the overhead, so the kp should be reduced accordingly when the ki increases. After adjusting kp and Ki, the effect is a certain amount of superadjustment, but the final result is still stable at the set value. Finally, we start to adjust Kd. The adjustment is still from small to large, and the adjustment of kd can minimize the amount of extra calls.

There is also a adjustment method for PID control algorithms --The Ziegler-nickles method.

The debugging method is to first set the integral and differential gain to 0, and then the proportional gain increases gradually from zero until it reaches the limit gain.KUAt this time, the controller output value oscillates with a constant value.KUAndOscillation cycleTUSet the ratio, integral, and differential gain according to different types in the following table.

Ziegler-nickls Method[2]
Control Type
Proportion		-	-
Proportion-Points			-
Classic ratio-integral-differential (PID)[3]
Pessen Integral Rule[3]
Some overshoot[3]

-- Wikipedia Co., http://zh.wikipedia.org/wiki/Ziegler%E2%80%93Nichols%E6%96%B9%E6%B3%95.

It seems much simpler than the first one. But now the question is what the Ku should be adjusted.

Self-irresponsible understanding: Ku should be set to adjust the kp value when ki and kd are all set to 0, so that the kp at this time is the ku when the controlled amount is fluctuating.

There are two other methods mentioned on the wiki. If you don't take a closer look, you have time to study them again.

At the end of this article, I just started to learn pid, which indicates that there must be many mistakes and omissions in this article. I hope you will correct your criticism.

In the next article (the last one in the project), let's give you the pid control algorithm code you have written ..