TABLE OF CONTENTS Because self-regulating and non-self-regulating processes have a different character, the tuning procedures that are applicable to one may not be applicable to the other. In this section, we will restrict our comments to self-regulating processes. The next section is devoted entirely to tuning controllers for non-self-regulating processes, primarily liquid-level control.
Most loops are tuned by an experimental technique. Even when a formal technique, such as open-loop or closed-loop testing, is used to determine the initial tuning, a final bit of fine tuning may be in order. Trial-and-error tuning requires the user to observe the response of the loop to a previous event, either a set point change or load change, then decide what tuning parameter (or parameters) should be changed, in which direction, and by how much. Experience helps in interpreting the response. The user must also have a thorough understanding of the effect of changing each of the tuning parameters.
For a self-regulating process controlled by a PI controller (interactive or noninteractive), the tuning map on Figure 6-2 depicts the response to a set point change for various combinations of proportional and integral tuning. The graph in the upper left-hand corner (graph Al) depicts the closed-loop response to a set point change when the controller is tuned with very low gain (wide proportional band) and no integral action. If integral action cannot be turned off, then the left-hand column represents minimum integral action, the largest possible value...
