8.6 IMC Background

The main advantage to IMC is that it provides a transparent framework for control-system design and tuning. As we show in this chapter, the IMC control structure can be formulated in the standard feedback control structure. For many processes, this standard feedback control structure will result in a PID controller (sometimes cascaded with a first-order lag). This is pleasing because we can use standard equipment and algorithms (i.e., PID controllers) to implement an "advanced" control concept.

The IMC design procedure is exactly that of the open-loop "control" design procedure developed in Section 8.3. Remember that a factorization of the process model was performed so that the resulting controller would be stable. If the controller is stable and the process is stable, then the overall controlled system is stable. This is true simply because if two transfer functions are stable, then the transfer functions cascaded together (multiplied) are stable. This is a nice result because in a standard feedback control formulation, the controller and the process can each be stable, yet the feedback system may be unstable. The restriction of this design procedure is that the process must be stable. An extension to unstable processes is covered in the Chapter 9.

Although the IMC design procedure is identical to the open-loop "control" design procedure, the implementation of IMC results in a feedback system. Thus, IMC is able to compensate for disturbances and model uncertainty, while open-loop "control" is not. Note that the internal model controller must be detuned to assure stability if there is model uncertainty.