(6c) Importance of Regulatory Control Design for MPC

Greg Shinskey has had a tremendous impact on the practice of process control engineering going back to the late 1970s and 1980s. During this time frame, the arrival of the Distributed Control System (DCS) and the minicomputer allowed Shinskey’s process-oriented, advanced regulatory controls to be readily implemented. Taking further advantage of increasing computing power, Model Predictive Control applications came on the industrial scene in the mid 1980s, with significant growth coming in the 1990’s and thereafter. MPC provided built-in capability to handle process/control interactions, constraints, and optimization, ostensibly removing the need to improve or modify the regulatory design.

With the benefit of experience gained over many MPC applications, the recognition of the importance of the regulatory control design for MPC has only increased, and an acknowledged best practice is to jointly design both the MPC and regulatory controls. Design considerations include disturbance rejection, constraints, and optimization objectives. As a result, many of the techniques and reasoning behind Shinskey’s approaches to process control hold true today.

In this presentation, we will highlight, via examples primarily involving distillation columns, the considerations that go into the design of the regulatory controls and MPC, while making reference to Shinskey’s contributions.