Regarding PID control
Thread Content
PID control is the basic unit of system control as well as the fundamental element in entire engineering projects. Statistics show that 70% of control loops in use today are still single-loop controls. However, when it comes to optimizing PID control and tuning various parameters for target variables such as pressure, temperature, flow rate, and liquid level, many people remain at a superficial level. As for the principles behind PID, many also only have a surface understanding, knowing what it does without understanding why it works that way. This discussion is held in response to the initiative of HaiChuan’s leadership, with the aim of providing basic knowledge, helping people understand relevant principles, improving their skills, and fostering more professionals skilled in both basic and advanced control techniques, so as to promote the common prosperity of HaiChuan and bring benefits to society! We welcome everyone to participate actively; with your involvement, the quality of this forum will improve significantly. Thank you! ! !
I’ve adjusted the PID quite a few times before; it’s mainly done by looking at the real-time curve. In practice, temperature control valves are rarely set to automatic mode, or converted into flow control valves with relevant parameters. Generally, there isn’t much that can be adjusted in the song of flow control valves. Therefore, the main adjustments are made to the pressure-controlled valve and the hydraulic-controlled valve. ——It also depends on whether stability of pressure and liquid level is more important, or stability of flow rate is more important.
To adjust PID parameters, one must first understand the mathematical model underlying them; the key is to accumulate experience and draw conclusions through actual on-site adjustments, applying what has been learned to similar situations (variations may exist among different manufacturers, but the general approach remains the same). Eventually, one will be able to master the process of adjusting these parameters.
P is what we refer to as the degree of sensitivity, that is, the speed of response to changes in the measured signal. I is the integration time, that is, the length of time it takes to eliminate the deviation. D is the differential time, that is, the amount of lead time when the deviation changes.
As someone put it vividly, P is the proportional band or gain factor, which represents the present moment. I is the integration time, that is, the duration required to eliminate the deviation, and it refers to the past. D is the differential time, that is, the amount of lead time when the deviation changes, and it represents the future. Since PID integration takes into account the present, past, and future, it is a paradigm of classical control; during the startup of a production process, adjustments are made to the control parameters.
I have been using an increased integration time to eliminate the impact of the \"integration\" term on parameter adjustment. To put it simply, if the controlled parameter fluctuates in a regular pattern with each cycle lasting 30 minutes, I set the integration time to 30 minutes, and then this integration term loses its effect. I’m not sure if this understanding is correct
To eliminate the integral term, the parameter must be increased to infinity; your understanding is incorrect.
Besides the trial-and-error approach based on experience, are there any other reliable methods for tuning PID parameters in practical applications?
Sometimes, the integral rate is used to indicate the strength of the integration effect; the integral rate is the reciprocal of the integration time. Therefore, increasing the integral rate enhances the integration effect, which is equivalent to reducing the integration time in order to increase that effect.