PID regulation in the follow-up system - Database & Sql Blog Articles

Probe current voltage pin 420*4450 head diameter 5.0 over current current and voltage pin

GW MB10F 0.8A 1000V rectifier bridge

Supply 1206 blue light 1206 blue LED quality assurance original authentic large number

I thought that I had studied computer science and I was looking for a software development job. I remember that the most popular software development in the database was that I went to the computer research institute for an interview. Did people ask me if I would PB? I said that it will only be VB, and then people will tell me what kind of foundation is needed to get the software. . . . Then I will go home and wait. Later, it was a coincidence that it embarked on the road of automation control.

Recalling the experience of more than ten years of experience, because I am not a professional in automation control, I have encountered countless problems (as opposed to the time). Fortunately, there are no things that engineers can’t solve, and they are experiencing setbacks every time. After successful repeated training, the control requirements were finally realized. Let me talk about my PID adjustment experience in the follow-up system!

The control system requires the function of PID adjustment. The PID setting value is a random amount of real-time change (from the instantaneous value of the punching flowmeter), and the required output (the feeding amount of the loss-in-weight feeder) can be timely, Accurately follow changes in setpoints to make fast, accurate responses.

Just getting started with PID adjustment can be described as a fog, only knowing that P is the ratio, that is, the input deviation is multiplied by a coefficient; I is the integral, which is the integral operation of the input deviation; D is the differential, and the input deviation is differentiated.

Specifically, I don’t know how to use it. I didn’t know how to set the parameters. So I ran to the bookstore and opened a book on the principle of automatic control. I looked at the formula of the book and I felt the pressure was huge. I didn’t understand it. It's back.

At that time, I had the system manual of this S7-200. There was a PID program and related explanations. I read it carefully several times. I understood it a little bit, and downloaded the description of the FB41 function block of S7-300. I don't understand, no matter what, write the program first. Soon the equipment was assembled and could not be sent to the customer, because my program has not been debugged, so I started working on the commissioning work.

System shock --

In order to facilitate debugging, I first set the set value to a constant value for debugging. I didn't expect to encounter a problem at the beginning. After the equipment was running, I saw that the inverter panel changed the number back and forth between 0-50HZ, and the motor followed. The gasping of the system, the system oscillated, and the parameters were changed several times. The debugging was in a stalemate at the beginning. (At the time, it was not clear that the P value should be given to the small point first. The integration time should be given to the long point first. At the beginning, P and I should not be adjusted at the same time. The P setting is almost the same. Then I set the I, and I think of a parameter and try a parameter. The result is tossing. One day is still the same. I can't sleep in bed at night, thinking incessantly, where should I start? I obviously didn't find a trick.

Early the next morning, sitting in front of the computer, while checking the information (the company can already access the Internet via ADSL, although the speed is a bit slow), while sorting out the ideas, how can we not overshoot? Since it is overshoot, it means that the adjustment is too big. How can we adjust it one by one? Yes, first change the P value to a small point, then drop it to half, a little effect, and then try to reduce it by half. At this time, the P value is reduced by 10 times, the device starts, and the frequency rushes to the given value. Nearby, not overshooting, looking for a feeling (I am secretly happy), and then slightly change the P value to a larger point, after several adjustments, the P value is determined, you can feel that the value of P is too small, you can When you are bigger, it will fluctuate again.

Try adjusting the points again, and the points have a smashing effect. Is it related to time? With doubts, I started the setting of I value again. According to the experience of P value adjustment, I can't wait to reduce the I value by 10 times. As a result, the system started the shock of yesterday, what? This is how it happened. I tried it a few times and found out the law. The smaller the I value, the more powerful the adjustment. Haha! Finally found a little doorway. After many adjustments, the system was finally controlled by me, but it took more than 20 seconds to adjust, and I started to get angry again.

Adjustment cycle --

Where can I adjust? After thinking about it for a few days, I still had no results. I accidentally put FB41 in OB1. The miracle happened. It was transferred in a short time. I read the description of FB41 carefully. At that time, those charts and instructions were simply suffering. As far as the parameter table can be understood, some parameters are not understood. After combing a parameter and a parameter, only the parameter of the sampling period is written, and the value of the sampling time is consistent with the set value of OB35, so it is opened. In the property setting dialog box of the CPU, the time of viewing OB35 is 100MS, which is obviously much slower than the scanning time of OB1. The OB35 time is adjusted to 10MS, and the experimental result is much better than the original.

Multi-parameter adjustment --

Solved the problem of the previous few days, the system commissioning began the key link, simulate the change of the set value, look at the speed of the step response, and then measure the amount of feed at this time to see if the design requirements can be met, the result data Most of them are beyond the scope of precision requirements, or not adjusted well, but the place where the adjustment is adjusted, I am a little depressed, is this effect, I am not willing to change, different parameters, will experiment The data were recorded one by one. After careful analysis, it was found that the small parameters (sm value is small, I value is certain) in the steady state, the data accuracy is relatively high, but the dynamic following is slow; the large parameter (the P value is large, the I value is constant) The step response is very good, but the data error is large; real-time and stability are mutually constrained. How can we do both fish and bear's paw? I have a chance to use the two sets of parameters when adjusting the MM440 inverter. Why not design two sets of parameters, set a deviation band according to the deviation between the set value and the feedback value. A set of parameters with a strong effect, once adjusted to the vicinity of the set value (within the deviation band), switches to a set of parameters with weaker regulation, avoiding overshoot and causing system oscillation. After multiple sets of parameters, the feeding accuracy of my experiment can meet the requirements, and the performance of automatic follow-up is also controlled within 3 seconds.

Undisturbed switching --

The equipment needs to be added to the silo every time. Because of the weight loss feeding method, the data in the feedback queue during the feeding is inaccurate. At this time, the PID function cannot be used, and it needs to be switched to the follow-up open loop mode. It is easy to do, the MAN value is always calculated according to the flow meter (always consistent with SP_INT), directly set LMN_ON to 1, you can cut it to manual, you can manually switch back to automatic, you must pay attention to it, this should be grasped Two points: 1. The PID feedback value (PV_IN) has been detected as normal. 2. Give the feedback value (PV_IN) at this time to SP_INT, and then return to normal adjustment. In fact, the PID of the follow-up system has a certain anti-disturbance capability.

Section --

Through the PID adjustment of this follow-up system, I feel that it is very interesting to do the PID adjustment control. It is like having a certain intelligence, and the feedback signal is detected at any time. Once the feedback value is found to deviate from the given value, it is given. The corresponding control force, even if the interference exists at any time, can eliminate the interference, because the controller already has a countermeasure strategy.