摘要：The variability within a process can be seen as a limiting factor for its efficiency. Because this, the present paper provides a simple, fast and non-intrusive methodology, able to translate the part in total variability that can be changed through controller adjustment. These types of analyses are usually intrusive; however, this work presents an alternative to obtaining such information from non-intrusive manner, thus eliminating a need for system perturbation. It is proposed to estimate the variability reduction through a sequential method that uses unmeasured disturbance estimation and a representative model of the actual process. The method’s quality was tested by analyzing the sensitivity of methodology against several initial conditions, since the procedure requires an optimization step. The case studies that involved pure time delay have showed high sensitivity to the initial condition, but even in this case the potential for variability reduction was correctly estimated.
