Robust distributed control for plantwide processes based on dissipativity in quadratic differential forms

Abstract

Based on dissipativity theory, a robust distributed control approach for plantwide chemical processes are developed in this work. The plantwide process is represented as a network of process units and (distributed) controllers. The model of each process unit with uncertainty is represented as a polytopic set. The effects of uncertainties of process units on plantwide stability are analyzed based on the concept of dissipativity. As a result, the plantwide robust stability condition is formulated in terms of dissipativity, which is conveniently represented as the linear combination of the dissipativity conditions of individual process units and controllers, based on the topologies of the process and controller networks. Dissipativity conditions (storage functions and supply rates) in the Quadratic Differential Forms are adopted to reduce the conservativeness of the proposed robust control approach. Furthermore, the dissipativity conditions that individual controllers need to satisfy to ensure plantwide robust stability is developed. A robust distributed control synthesis approach for designing individual controllers is also developed.