Analog and digital PID controllers

The standard controller most widely used in industry is the PID (proportional-integral-derivative) controller, because its three parameters allow you to set the performance (damping, response time) of a process control modeled by a second order. There are many physical systems which, even if complex, have a behavior close to that of a second order, within a certain time scale. As a result, the PID controller is well suited to most industrial-type processes, and is relatively robust to variations in process parameters, provided that the performance of the closed loop is not too demanding compared to that of the open loop (e.g., very significant acceleration of response or very significant increase in closed-loop damping).

If the dominant dynamics of the system is greater than second order, or if the system contains a large delay or several oscillating modes, the PID controller is no longer adequate and a more complex controller (with more parameters) must be used, at the expense of sensitivity to process parameter variations.

The implementation of a PID control loop is a very important problem, because it influences :

• the quality of regulation on an industrial site ;

• the time required to implement the order ;

and has two essential aspects:

• PID controller tuning, for which knowledge of a dynamic process model on the one hand, and desired performance on the other, determine the choice of synthesis method;

• implementation of the controller in analog or digital version, and in serial, parallel or mixed configuration.

Increasingly, PID controllers on the market offer the option of auto-tuning, which automatically calculates parameters at the user's request.

Paragraphs 1 , 2