PID control to play and see

PID control to play and see

PID three parameters exactly how to adjust, really is an art.
Although there are self-measured functions in the Marlin firmware, the results are not as satisfying as they are, or need to be adjusted. But how exactly should it be adjusted? from which parameters do you start? Does the value increase or decrease? will not adjust this, but also need to go back to adjust the last one? The whole adjustment of the Cheng in the chaos is seven or eight bad, not easy to get out of the head. It takes time to add an experiment, but it's tiring to find the best parameters.
Today, I don't hyo what is wrong, I suddenly want to say that you can use the Excel simulation PID temperature control, so you can quickly get the experiment results, increase the tuning PID parameters of the experience. Although there is no way to directly simulate the PID parameters that are needed to crowded out the heat, you can at least be familiar with tuning PID parameters for temperature control.
When you download Excel and open it, you can see that the upper-left corner has an adjustable number of parameters, the middle is the simulation process, and the right side has two simulation results. The simulation parameters are divided into two groups, the first of which are the environment variables, and the last three are the PID values.
be hotter than hot: The number of calories per g/, the amount of calories required, can rise by 1 ℃. The heat of the water is 1 (Card/g/-℃)
The amount of heat subject matter: (It doesn't have to be explained ...)
Initial temperature: (It doesn't have to be explained ...)
temperature of the environment: (It doesn't have to be explained ...)
temperature of the target: (It doesn't have to be explained ...)
Heat Dissipation Constants: constant associated with the heat dissipation rate. The system loses the heat, I use this side (at the temperature-environment temperature) x thermal constants to simulate the calculations. The larger the heat dissipation constant, the faster the system loses its heat, and the greater the temperature difference between the system temperature and the environment, the faster the heat dissipation is.
PID: The proportional control constants, the machine control constants, the differential control constants.
System Heat Supply: This is the percentage of power that is applied to the hot rod. The simulation formula for this "system heat supply" is as follows:
"This time" system heat supply
P x this time to take the same temperature error +
I x from the system start to take the same temperature error every time and +
D x (the temperature error of the last time-this time the temperature error)
P: Proportional control constants. The error value is multiplied by the P value, and the system heat supply is added back.
If only P,i and D were set at 0, I thought that when I could compare him to a car, the gate would only "directly" step on a certain value, or completely release it. Does not slightly adjust the force of the oil door. In theory, only proportional control can still allow the temperature to be controlled near the target temperature, but the oscillations will be more severe. Just like the car, always slam on the oil door, and slam the brakes, the car will run, the speed also has, but passengers would dizzy to death ...

D: Differential control constants. Temperature error The curve of the time is a two-dimensional curve. And the meaning of the differential of the two-dimensional curve is the slope. So the differential control actually takes the last time to subtract from this same temperature error, dividing it by two times, and calculating the slope. Multiply the slope by D, and then add the system heat supply.
The effect of the D-value, I would liken him to an "early rate decrease". When the speed of the car is close to the speed at which the machine wants to hold it, the experience of the machine will start to reduce the oil door first. Do not wait for the speed of the head to the brakes, reducing the speed of the car to reduce the vibration, so that passengers comfortable not dizzy car. How many times in advance to reduce speed, how much speed should be reduced, or minus too little, depends on the size of the D value to adjust.

The addition of D "differential control" allows the system to be more stable to the target temperature.

A suitable D-value allows the temperature to be quickly and steadily fixed at the target temperature, and there is no problem.

D value is too large, the early braking force is much more, but it will make the system unstable and vibration.

Continue to increase the D value, and even let the results of control can not be collected 斂 to the target temperature.

I: Integral control constants. The concept of an integral part of this is to sum up all the mistakes of the past. The control of the integral is, of course, the sum of the control constants multiplied by all the misunderstandings, plus the system heat supply.
The integral control is used more to correct the stability error. When the temperature is stable, the temperature is still slightly wrong, which is the "stability error." If the P-value is adjusted for stability, the stability of the whole system will be greatly affected. The integral control relies on the accumulated stability error, so that the error of the effect due to the accumulation and amplification, and then multiplied by the machine control constant, you can eliminate Praya stability error.
These are some of my thoughts on PID control. Special attention to my introduction of the time, the order is PDI, the intention is to recommend the adjustment of the time, you can first back to adjust P, D these two constants (I return 0), feel the system is stable, satisfied, and then the "Stability error" to adjust the I-integral control constants.
Modifying the Mass and heat of the subject, you can observe the influence of the thermal system on the PID control. Modify the thermal constants so that you can observe the system's thermal situation (with more fan-blown pipes), which affects the system's stability. Through the simulation experiment, many more environment, and adjust the relative PID value, should be more familiar with PID control this art ~

References

About PID knowledge and discussion?

PID control to play and see