11.1 Background

The focus of the techniques developed thus far in this textbook has been on linear systems, although it is recognized that chemical processes are inherently nonlinear. Fortunately, when processes are operated close to a desired operating point, the linear-system approaches generally work well. Also, when the main nonlinearity is due to the control valve or flow-pressure drop characteristics, cascade-control approaches work quite well. The inner-loop flow controller has fast dynamics, and the nonlinear effects will usually not propagate to the outer-loop.

In this chapter we cover several different approaches to handling nonlinear effects. The first nonlinear effect is due to manipulated input saturation. When manipulated inputs hit a constraint, the phenomena of reset (or integral) windup can occur. Approaches for antireset windup (ARW) are presented in Section 11.2.

When chemical processes change operating points, the dynamic input-output characteristics often change, requiring a control system retuning. Automatic tuning procedures can be used to retune controllers. So-called autotune procedures are developed in Section 11.3.

Sometimes it is desirable to have a controller gain that is a function of the error. If small errors can be tolerated, then a controller with a proportional gain that is a function of the absolute value of the error can often be used. The basic idea of nonlinear PID is introduced in Section 11.4. If the process gain changes as a function of operating condition, then "gain scheduling" approaches can be used to vary the controller tuning parameters as a function of the operating condition. Gain scheduling is presented in Section 11.5. Sometimes process input-output behavior can be made more linear by proper measurement/actuator selection. Such techniques are discussed in Section 11.6.