Understanding PID
Thread Content
Regarding PID, I have been reading some materials recently, and there are some aspects that I don’t quite understand. I hope that through this post, others can point out any misunderstandings I might have regarding PID control. Thank you! The larger the value of P, the smaller the steady-state error. However, once P reaches a certain level, oscillations can occur. The smaller the value of I, the stronger the integrating effect, which results in a faster response; conversely, a larger value of I leads to a slower response. A longer integration time increases the likelihood of overshoot. The function of integration is to accumulate errors over time and then output them all at once, so a longer integration time results in poorer system stability. Differentiation serves to respond in advance to overshoot, acting as a kind of lead time; however, if its value is too large, it can also cause oscillations. Is there anyone who can provide some detailed additional information on PID control? Don’t be too theoretical
Proportional action P is based on the deviation; when the deviation remains constant, the greater the proportional action, the stronger the control effect. Therefore, an excessive proportionality can lead to overshoot. A certain level of overshoot is necessary, usually around 2-5%, but if the proportionality is too high, it will result in overly strong control and thus oscillations. Proportional action is similar to walking: your brain constantly directs your feet to move forward, and when there is a deviation due to stepping off course, adjustments are made to correct it. However, if these adjustments are too large, it will cause erratic oscillations. Integral action serves as a further correction to proportional action, as proportional action alone cannot ensure that the actual value exactly matches the set value – in other words, a deviation always exists. When walking, if you notice that one step is longer than the desired distance, you will take a smaller step next time; if it’s still too long, you’ll take an even smaller step. As long as there is a deviation, your brain will actively adjust your steps to get closer to the target value, and this adjustment is calculated based on the cumulative effect of the deviation over time – this is what differential action does. Mathematically, integrating the deviation over time gives the cumulative deviation over time, and this value is proportional to the control amount, with the ratio being the integration time constant. If the deviation remains constant, the time required for the control amount to double is the integration time. Integral action involves making gradual adjustments as long as the deviation persists. Differential action is easy to understand as well; using walking as an example, if your leg tends to move left while the target is straight ahead, your brain will adjust your next step in advance to avoid deviation. Differential action is based on the trend and rate of change of the measured values, which means it can sometimes cause unnecessary reactions and lead to oscillations. It’s generally sufficient to use it only in situations with significant delays.
The greater the proportional action, the stronger the control effect. It seems more appropriate to refer to this proportion as gain; in PID calculations, the proportional term is actually taken into account as its reciprocal.
The poster’s explanation is more in line with the actual conditions at the production site.
Differential is for fast tracking; it is not used in most cases, and if it is used, its value is generally required to be small, otherwise oscillations are likely to occur. Integration actually reflects the lag time.
Specific operating conditions, circuit types, and the status of instruments and valves are all factors that need to be taken into account when adjusting parameters; even if a single parameter is set to automatic control, the results will not be very good~
I’m sending you an amazing article; once you read it, you’ll understand it even better