An evaluation of a hierarchical multi agent based process monitoring system for chemical plants

Multi agent systems have gained increasing interest in research, particularly in areas where distributed intelligence is required, in situations where single centralized methods are not suited. Fault detection and identification in chemical process industries is one such domain where multi agent systems would be ideally suited. A chemical process industry experiences various faults at different levels of granularity in the plant; some may be at the sectional level, equipment level or at the sensor level and so on. Such faults if allowed to propagate could result in leaks, fires and explosions, resulting in loss of life, capital invested and production downtime. Single centralized monolithic monitoring strategies are not suited for identifying such faults as deviations in the process are lagging indicators by which time plant safety may be compromised. One method of isolating such faults would be to look at the process hierarchically in order to identify sections or equipments which may be faulty and then in turn to focus on the particular section for isolating faults. Therefore, a hierarchical multi agent based process monitoring system is studied in this paper to evaluate its advantages and disadvantages. Several FDI agents are developed at different levels of granularity of the process. The agents are incorporated into a multi agent system called ENCORE (Natarajan and Srinivasan, 2010) developed in our prior work. The benefits vis-à-vis the drawbacks of using such a hierarchical methodology are presented with an application to a simple CSTR — Distillation Column case study.

[1]  S. Joe Qin,et al.  Reconstruction-Based Fault Identification Using a Combined Index , 2001 .

[2]  J. Edward Jackson,et al.  A User's Guide to Principal Components: Jackson/User's Guide to Principal Components , 2004 .

[3]  Raghunathan Rengaswamy,et al.  A review of process fault detection and diagnosis: Part II: Qualitative models and search strategies , 2003, Comput. Chem. Eng..

[4]  J. Edward Jackson,et al.  A User's Guide to Principal Components. , 1991 .

[6]  Rajagopalan Srinivasan,et al.  Multivariate Temporal Data Analysis Using Self-Organizing Maps. 2. Monitoring and Diagnosis of Multistate Operations , 2008 .

[7]  Yew Seng Ng,et al.  Evaluation of decision fusion strategies for effective collaboration among heterogeneous fault diagnostic methods , 2011, Comput. Chem. Eng..

[8]  J. E. Jackson.  A User's Guide to Principal Components , 1991 .

[9]  Rajagopalan Srinivasan,et al.  Multivariate Temporal Data Analysis Using Self-Organizing Maps. 1. Training Methodology for Effective Visualization of Multistate Operations , 2008 .

[10]  Ali Cinar,et al.  Statistical process monitoring and disturbance diagnosis in multivariable continuous processes , 1996 .

[11]  Raghunathan Rengaswamy,et al.  A review of process fault detection and diagnosis: Part I: Quantitative model-based methods , 2003, Comput. Chem. Eng..

[12]  Raghunathan Rengaswamy,et al.  A review of process fault detection and diagnosis: Part III: Process history based methods , 2003, Comput. Chem. Eng..

[13]  Venkat Venkatasubramanian,et al.  Challenges in the industrial applications of fault diagnostic systems , 2000 .

[14]  Edmund H. Durfee,et al.  Negotiating Task Decomposition and Allocation Using Partial Global Planning , 1989, Distributed Artificial Intelligence.