Simulating Industrial Alarm Systems by Extending the Public Model of A Vinyl Acetate Monomer Process

The research area of industrial alarm monitoring and management has been attracting increasing studies in the past decade. A variety of advanced alarm management techniques have been proposed. However, the validation of the results were mostly based on real industrial data while such data was not always accessible to all researchers. Even if obtained, the privacy policies of industrial corporations may lead to difficulty in publishing new results. Thus, there is a great demand to construct a large scale simulated alarm system and create standard alarm & event data for the test of new methods. This paper proposes the design of a simulated alarm system by extending the public model of a Vinyl Acetate Monomer (VAM) process and collects the alarm & event data through a long term alarm monitoring process. An example alarm & event dataset is created and analyzed to illustrate the usability of the simulated alarm system.