Understanding Process Control: Feedback, Feedforward, and Advanced Strategies
Understanding Process Control
The gain of the controller can be adjusted to make the output changes as sensitive as desired to deviations. The sign of the gain can be chosen to make the controller output increase or decrease as the error signal increases. The bias must be adjusted so that when there is no error (steady-state desired conditions), the controller output and the manipulated variable are at their nominal steady-state values. Whenever the controller gives a signal that makes the control equipment be at, or surpass, their physical bounds, we say that the controller is saturated. There is always offset when we change the set point, or the disturbances produce a sustained change in the system. In order to eliminate it, we have to manually reset the set point or the bias
Integral control action is widely used because it eliminates the offset. If errors cannot be eliminated quickly, integral control action produces larger and larger values, which keeps increasing the control action until it is saturated. This is called integral windup. The integral action tends to produce an oscillatory response of the controlled variable and reduces the stability of the feedback control system.
The controller anticipates what the error will be in the immediate future by considering the rate of change of the error signal. If there is a constant error, the derivative action will give no control action. If the process measurement is noisy, the derivative of the measured variable will change widely, and derivative action will amplify the noise. A sudden change in the set point changes the error abruptly, and the derivative term provides a derivative kick to the final control element. This can be avoided if, instead of the derivative of the error, we use the derivative of the process variable.
The closed-loop response of a PID has, in general, similar qualitative dynamic characteristics as the response of a PI controller – The derivative control action has a stabilizing effect on the system. However, it presents certain drawbacks when the system changes rapidly. Approximately 90% of all controllers in the industry are PI, and when this derivative action is needed, these are substituted by model predictive control.
Disadvantages of Feedback Control
- No corrective action is taken until the deviation in the controlled variable occurs. Therefore, perfect control is impossible.
- It does not provide predictive control action to compensate for known or measurable disturbances.
- It might not work properly for processes with large time constants or time delays since they may never attain a steady state and operate continuously in a transient state.
- It requires the controlled variable to be measured online, which cannot happen in some cases.
The basic concept of feedforward control is to measure the important disturbance variable and take corrective action before they modify the process.
Feedforward and Cascade Control
As mentioned before, a disadvantage of feedback control is that the corrective action does not start until after the controlled variable has deviated. Feedforward control might help with this, but then we need to measure the disturbances explicitly and have a model to calculate the controller output. An alternative approach, cascade control, employs a secondary measurement point and a secondary feedback controller. This secondary measurement point is located such that it can recognize the change in the conditions sooner than the controlled variable, without having to necessarily measure the disturbance.
Example: Cascade Control in Temperature Control
- Master loop: Temperature control (2 Feedback)
- Secondary control loop: Pressure control
If there is a disturbance in the supply pressure, the pressure controller will act quickly to maintain the pressure at the set point.
- The output signal of the master controller serves as the set point for the slave controller.
- Both feedback control loops are nested.
Ratio Control
Ratio control is a special type of feedforward control with widespread application in the industry. If we want to maintain the ratio of two process variables at a specified variable, we consider the ratio between the process variable Χ and a disturbance variable ξ, and control it instead.
Approaches to Ratio Control
- Direct control approach: It directly adjusts the manipulated stream based on the disturbance stream.
- Ratio station: Calculates a setpoint for the manipulated stream, and a separate flow controller adjusts the actual flow to match this setpoint.
- Makes process gain non-linear, worse than cascade control.
Split Range Control
Split range control has one measurement and more than one manipulated variable. Used when the action needed to manipulate a process can’t be achieved with a single control element or when the process needs different actions depending on the level of the measured variable. Benefits: Efficient energy use, smooth transition, simple control system.