A Framework for Constraint Classification and Management in Dynamic Environment of the Process Industry

Abstract

Vigorous research in the field of constraint programming has led to the expectation of handling various types of constraints in a dynamic environment such a chemical process industry where constraints tend to alter their scope of impact and sometimes even context in which they are used.  Constraint programming languages are developed to handle such scenarios focusing more on domain specific problem. These approaches ensure modeling in a particular domain but extending such models for other domains becomes unmanageable to suit needs. In this research, we propose a theoretical framework which gathers all types of constraints that are possible in a dynamic environment with an objective to organize and administer them. The aim is achieved by having functionalities for observing their scope by setting their preferences and priorities according to their lifespan in the system. The type of constraint determines the technique in which it is supposed to be handled. Since, the framework modules work in collaboration to help find a solution to the problem; we use multi-agent technology for catering environment dynamicity. The principle contribution can be considered in classifying constraints and managing them independent of their domain. For validation, a chemical continuous flow process scenario is studied as an application.