The general concept of cascade control is to place one feedback loop inside another. In effect, one takes the process being controlled and finds some intermediate variable within the process to use as the set point for the main loop.
Cascade Control Loops
Cascade control exhibits its real value when a very slow process is being controlled. In such circumstances, errors can exist for a very long time, and when disturbances enter the process there may be a significant wait before any corrective action is initiated.
Also, once corrective action is taken one may have to wait a long time for results. Cascade control allows the operator to find intermediate controlled variables and to take corrective action on disturbances more promptly.
In general, cascade control offers significant advantages and is one of the most underutilized feedback control techniques.
An important question you may confront when implementing cascade control is how to select the most advantageous secondary controlled variable.
Quite often, the designer has a large number of choices. The overall strategy or goal should be to get as much of the process lag into the outer loop as possible while, at the same time, getting as many of the disturbances as possible to enter the inner loop.
The next two figures illustrate two different cascade control arrangements for a furnace that is used to increase the temperature of a fluid that is passing through it.
In both cases, the primary controlled variable is the same, but in each case a different intermediate controlled variable has been selected. The question then is which type of cascade control is best.
To determine the best cascade control arrangement, you must identify the most likely disturbances to the system. It is helpful to make a list of these in order of increasing importance.
Once this has been done, the designer can review the various cascade control options available and determine which one best meets the overall strategy outlined earlier: to make the inner loop as fast as possible while at the same time receiving the bulk of the important disturbances.
If both controllers of a cascade control system are three-mode controllers, there are a total of six tuning adjustments.
It is doubtful that such a system could ever be tuned effectively. Therefore, you should select with care the modes to be included in both the primary and secondary controllers of a cascade arrangement.
For the secondary (inner or slave) controller, it is standard practice to include the proportional mode. There is little need to include the reset mode to eliminate offset because the set point for the inner controller will be reset continuously by the outer or master controller.
For the outer loop, the controller should contain the proportional mode. If the loop is sufficiently important to merit cascade control then you should probably include reset to eliminate offset in the outer loop.
You should undertake rate or derivative control in either loop only if it has a very large amount of lag.
The tuning of cascade controllers is the same as the tuning of all feedback controllers, but the loop must be tuned from the inside out.
The master controller should be put on manual (i.e., the loop broken), and then the inner loop can be tuned.
Once the inner loop is properly tuned, the outer loop can be tuned. This allows the outer loop to “see” the tuned inner loop functioning as part of the total process or as the “all else” that is being controlled by the master controller.
If you follow this general inside-first principle when tuning cascade controllers, you should encounter no special problems…