Integrated Modelling of Process Operation Systems Using the Agent-Oriented Approach

For the past years, several software and computer tools have been developed to aid the chemical process operations including real-time simulation, on-line optimization, fault diagnosis, process monitoring, and many other functions. These tools were designed separately and did not collaborate efficiently, making it difficult to integrate different engineering tasks for the optimal process operation. In this paper, an agent-oriented modelling approach is presented to address this problem. Elements in the process operation systems are divided into two classes. One class consists of equipment, units and processes, while the other class consists of production operation tasks. The two classes of elements are modelled as objects and agents, respectively. Then, three strategies are presented to implement the integration of the whole process operation system, which are integration of object models, integration of agent models and supervision of operator. Also presented is a case study of integration of process operation decision optimization and abnormal situation management using the proposed agent oriented approach for TE challenge problem.