A direct control parameter tuning method using generalized minimum variance evaluation

In process control, a key of acquiring desired control performance is making good use of information that is included in operation data. Direct controller adjustment methods without the knowledge of a process model, and techniques for diagnosing control performance, which is control performance monitoring / assessment (CPM / CPA) have been proposed. The present work proposes a direct control parameter tuning method based on generalized minimum variance (GMV) evaluation for regulatory control. The proposed method derives control parameters that can minimize the variance of estimated the generalized output which is generated from a set of closed-loop experimental data. Moreover, control performance of the proposed method can be evaluated by GMV based index. The efficiency of the proposed method was demonstrated through simulations.