To use the PID adjustment, it is necessary to understand the meaning of the PID calculation formula and PID parameters. Here is the formula for PID: Suizhou simi intelligent technology development co., LTD , https://www.msmvape.com
Where the error signal e(t)=SP(t)–PV(t), M(t) is the output value of the PID controller, Kc is the gain of the controller (proportional coefficient), Ti and Td are the integration time and The derivative time, Minitial is the initial value of M(t), which is actually the initial value of the integral.
The first three terms of the PID formula are proportional to the error, the integral of the error, and the derivative of the error.
Differential and integral are the concepts of advanced mathematics. It is suggested that netizens who have not studied advanced mathematics should at least understand the geometric meaning of differential and integral, which is of great help to understand the meaning of PID parameters.
The integral corresponds to the error curve e(t) in the figure below and the area enclosed by the coordinate axis (the gray part in the figure). The PID program is executed periodically, and the time interval for executing the PID program is Ts (ie, the sampling period controlled by the PID). We can only use the values ​​of some discrete points on the continuous error curve with an interval of Ts to calculate the integral, so it is impossible to calculate the exact integral value, and only the integral can be approximated.
The sum of the rectangular areas in the figure below is generally used to approximate the exact integral. When Ts is small, the error of the integral is not large.
Make a tangent to the error curve e(t) (see the following figure). The tangent value tgα of the angle α between the tangent and the positive direction of the x-axis is the first derivative de(t)/dt of the error at that point. The derivative in the PID controller output expression is approximated by:
De(t)/dt≈Δe(t)/Δt=[e(n)-e(n-1)]/Ts, where e(n) is the error value at the nth sampling, e(n- 1) is the error value at the n-1th sampling.
PID regulation is currently the most widely used regulation control law, P proportional, I integral, D differential control, referred to as PID control.
Proportional control is one of the simplest ways to control. The proportional effect is large, which can speed up the adjustment and reduce the error, but the excessive proportion makes the stability of the system drop, and even causes the instability of the system.
Integral adjustment allows the system to eliminate steady state errors. If the system has a steady-state error after entering the steady state, it is necessary to introduce an "integral term". Proportional + integral (PI) control allows the system to have no steady-state error after entering steady state.
The differential action produces an advanced control that has been eliminated by differential regulation before the deviation has not yet formed. Therefore, the dynamic performance of the system can be improved. . For controlled objects with large inertia or hysteresis, proportional + derivative (PD) control can improve the dynamic characteristics of the system during the adjustment process.
This is an excerpt of a PID parameter adjustment for everyone to learn reference:
Parameter tuning finds the best, from small to large, the first step is the proportional and post-integration. Finally, the differential plus curve is oscillated very frequently. The proportional disk needs to be enlarged to float around the big bay, and the proportional disk is slower than the return curve. The fluctuation period of the downward curve is long, and the integration time is longer. The oscillation frequency of the curve is fast. The differential is first lowered and the fluctuation is large and the fluctuation is slow. The differential time should lengthen the two curves of the ideal curve, 4 to 1 before the high and then lower
When you look at the two-tone analysis, the quality of the adjustment will not be low.
This jingle spreads widely, I think the operability is very low (may also be my lack of understanding), I have a lot of questions:
"From small to large, check", what to check?
Must be "first proportional and then integral"? Is it not good to use PI directly?
“The curve oscillates very frequentlyâ€, does it mean that the oscillation frequency is high or the number of oscillations is high?
What is a "proportional disk"?
"What does the curve float around Dawan?" Does it mean that the amount of overshoot is large? Still rising slowly?
Is the period of "long curve fluctuation period" the oscillation period? Or is it excessively long?
Is the oscillation frequency and differential relationship large? What is the main role of differentiation?
"Two waves of ideal curve", is a wave 180 degrees or 360 degrees? The two waves are ideal curves. The PV curve in the figure below is not ideal.
I have used the PID adjustment control panel and PID parameter auto-tuning function of S7-200 and S7-200SMART, and the controlled object is simulated by the subroutine I wrote. The parameters of the controlled object are as follows: the gain is 3.0, and the time constants of the two inertia links are 5s and 2s.
The following is the curve before auto-tuning, the overshoot is too large:
The following is the curve of the tuning process:
The following is a curve of the parameters obtained by tuning:
Below is another set of curves for the parameters before tuning. The process variable PV curve rises too slowly:
Although the PID parameters of the two groups before the tuning are very different, the PID parameters are almost the same after two tunings, and the shape of the PID parameters obtained by the tuning is similar.
I think Siemens' PID parameter auto-tuning is very useful.
To use the PID adjustment, it is necessary to understand the meaning of the PID calculation formula and PID parameters. Here is the formula for PID:
Where the error signal e(t)=SP(t)–PV(t), M(t) is the output value of the PID controller, Kc is the gain of the controller (proportional coefficient), Ti and Td are the integration time and The derivative time, Minitial is the initial value of M(t), which is actually the initial value of the integral.
The first three terms of the PID formula are proportional to the error, the integral of the error, and the derivative of the error.
Differential and integral are the concepts of advanced mathematics. It is suggested that netizens who have not studied advanced mathematics should at least understand the geometric meaning of differential and integral, which is of great help to understand the meaning of PID parameters.
The integral corresponds to the error curve e(t) in the figure below and the area enclosed by the coordinate axis (the gray part in the figure). The PID program is executed periodically, and the time interval for executing the PID program is Ts (ie, the sampling period controlled by the PID). We can only use the values ​​of some discrete points on the continuous error curve with an interval of Ts to calculate the integral, so it is impossible to calculate the exact integral value, and only the integral can be approximated.
The sum of the rectangular areas in the figure below is generally used to approximate the exact integral. When Ts is small, the error of the integral is not large.
Make a tangent to the error curve e(t) (see the following figure). The tangent value tgα of the angle α between the tangent and the positive direction of the x-axis is the first derivative de(t)/dt of the error at that point. The derivative in the PID controller output expression is approximated by:
De(t)/dt≈Δe(t)/Δt=[e(n)-e(n-1)]/Ts, where e(n) is the error value at the nth sampling, e(n- 1) is the error value at the n-1th sampling.
1. The key point of fuzzy control is to sum up a large amount of practical data, and then make a black box, seemingly mysterious, actually the empirical parameters!
2. The data obtained by fuzzy control is based on the fact that the performance of the control device is unchanged, which is more accurate. Once the usage time is long and the performance is degraded, these empirical parameters tend to be very biased.
3. Even if different equipments of the same type are different in process environment, process flow and process characteristics, their performance will be different, so the repetitive use of the same data in fuzzy control cannot be achieved.
4. The concept of fuzzy control is very good, at least it is advanced control, but for now, its practicality, dynamics is still not as good as the traditional PID.
5. The traditional PID is hysteresis control, and in most of the current process environments, it can still meet the control requirements.
6. Based on the characteristics of traditional PID, different control methods have also been extended, such as cascade adjustment, three impulse adjustment, split adjustment, stepping and so on. #p#分页头#e#
7. Personally think: With the rapid development of electronics, network and computer, the lag of traditional PID will be better, and its dynamic adjustment characteristics are unmatched by fuzzy control.
It is important to understand the physical meaning of PID parameters and the relationship between PID parameters and closed-loop system performance indicators.
The control principle of PID can be understood by manual control of the furnace temperature. First look at the role of the proportional part.
It is important to understand the physical meaning of PID parameters and the relationship between PID parameters and closed-loop system performance indicators. First look at the role of the proportional part.
The control principle of PID can be understood by manual control of the furnace temperature. The operator reads the measured value of the furnace temperature detected by the digital meter with the eye and compares it with the set value of the furnace temperature to obtain the temperature error value. Operate the potentiometer by hand to adjust the heating current so that the furnace temperature remains near the set value.
The operator knows the position of the potentiometer when the furnace temperature is stabilized at the set value (we call it position L) and adjusts the corner angle of the potentiometer based on the current temperature error value. When the furnace temperature is lower than the set value, the rotation angle of the potentiometer is increased clockwise on the basis of the position L to increase the heating current; when the furnace temperature is greater than the set value, the potentiometer is counterclockwisely reduced on the basis of the position L. The corners to reduce the heating current. The difference between the adjusted potentiometer rotation angle and the position L is proportional to the error. The larger the absolute value of the error, the larger the adjustment angle. The above control strategy is proportional control.
There are various delays in the closed loop. After adjusting the potentiometer corner, there is a large delay when the temperature rises to the steady state value corresponding to the new corner. Due to the delay factor, the adjustment effect cannot be seen immediately after adjusting the potentiometer rotation angle. Therefore, the main reason for the difficulty in adjusting the closed-loop control system is the delay effect in the system.
If the gain is too small, the adjustment is not strong enough to make the temperature change slow and the adjustment time is too long. If the gain is too large, the adjustment force is too strong, causing the adjustment too much, which may cause the temperature to fluctuate and fluctuate.
If the closed-loop system has no integral action, the simple proportional control has a steady-state error, and the steady-state error is inversely proportional to the gain. The larger the gain, the smaller the steady-state error, but the overshoot will increase, the number of oscillations will increase, and the closed-loop system will be unstable. Therefore, pure proportional control is difficult to balance dynamic performance with static performance.
