A comparison of two feedforward control structure assessment methods

In English The most common loop in an industrial plant is a SISO-loop with a PI(D)controller. If the loop is affected by disturbances,these can be handled by adding a feedforward control action. In an industrial control system there are many measured signals, apart from those of the loop, available. The question we address in this article is how to select the feedforward variable out of the available measurements. A comparison is made of two methods for feedforward control structure assessment. The first is a method based on comparing the actual variance with the minimum achievable variance,and thus the contribution to the overall variance. The second is a method for evaluating deterministic additive disturbances and estimate where they enter in the process. The methods complement each other and show that assessmentmethods should be handled with care, since they are mostly designed fordifferent control objectives.