PID adjustment

Adjust the secondary circuit first, then the primary circuit; once both are stable, activate the cascade control

The former is proportional-integral control, while the latter is pure differential control. (1) Proportional action establishes a proportional relationship between the controller’s input and output; in order to minimize the error as well as to speed up the response time and reduce the adjustment period, it is necessary to increase Kp. However, an excessive proportional action will degrade the dynamic performance of the system and may even render the closed-loop system unstable. (2) The introduction of integral action helps to eliminate steady-state error, but reduces the stability of the system. Especially during the large-deviation phase, integration often causes excessive overshoot in the system, prolonging the settling time. (3) The introduction of differential action enables the system to respond to the trend of deviation changes; an appropriate amount of differential action can accelerate the system’s response, effectively reduce overshoot, improve the system’s dynamic characteristics, and enhance its stability. The disadvantage is that the differential action is sensitive to interference, reducing the system’s ability to suppress it. Therefore, the parameter selection of a PID controller must take into account both dynamic and static performance requirements; only by properly setting the three parameters Kp, Ki, and Kd can satisfactory control performance be achieved.

Please prove it – are you referring to the two pictures posted by the original poster? If so, the second image still requires proportional-integral control, because D is 0. PS: In your setting, does the ‘P’ refer to proportional gain or proportional time?

Hey, original poster: if you can adjust this PID so that the sawtooth wave becomes a smooth wave, then I’ll stream myself eating shit and cheer for you – how about that? :P.

In Figure 2, the value of I is clearly too small, resulting in excessive overshoot. If the value of P in Figure 1 meets your requirements, there’s no need for significant adjustments. If neither the PI value in Figure 1 nor anything else meets the requirements, try increasing those values slightly; don’t make too many changes at once

That’s way too aggressive; just add a filter to smooth it out.

:Lol :lol Filtering, you little guy Cao Shujian – you really do know how to apply “filtering” to PID control signals, don’t you? Or can you actually apply “filtering” to the amount of movement of the device? If you’re so good, then teach the original poster how to do it properly! Go and smooth out that measured curve! Once you’ve done that, I’ll stream live as I eat a bowl of shit; if you produce a bowl of shit, I’ll eat it. How about that? :lol . . .

:Lol, little guy Cao Shujian – the opportunity to prove your skills is right in front of you. If you’re capable, you can help the original poster right now by coming up with your own \"filtering\" method to smooth out that sawtooth wave produced by the PID control, so that the controlled variable doesn’t fluctuate wildly. If you’re not capable, then your so-called \"filtering\" methods are nothing but useless nonsense, right? Little good boy Cao Shu Jian, hehe: lol hehe, be good, right? :P