Premium
Model‐predictive safety optimal actions to detect and handle process operation hazards
Author(s) -
Soroush Masoud,
Masooleh Leila Samandari,
Seider Warren D.,
Oktem Ulku,
Arbogast Jeffrey E.
Publication year - 2020
Publication title -
aiche journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.958
H-Index - 167
eISSN - 1547-5905
pISSN - 0001-1541
DOI - 10.1002/aic.16932
Subject(s) - process (computing) , particle swarm optimization , mathematical optimization , computer science , process safety , model predictive control , work in process , engineering , algorithm , mathematics , artificial intelligence , control (management) , operations management , operating system
In 2016, we introduced the concept of model‐predictive safety (MPS; Ahooyi et al, AIChE J . 2016; 62:2024‐2042). MPS is a proposed innovation in functional safety systems to methodically account for process nonlinearities and variable interactions to enable predictive, prescriptive actions, while existing functional safety systems generally react when individual process variables exceed thresholds. MPS systematically utilizes a dynamic process model to detect imminent and potential future operation hazards in real time and to take optimal preventive and mitigative actions proactively. This work expands the concept of MPS and formulates two min–max optimization problems, offline solutions of which are the optimal proactive preventive and mitigating actions that MPS takes online, in response to predicted process operation hazards. A nested particle‐swarm optimization (PSO) algorithm is proposed to solve the min–max optimization problems. The application and performance of the min–max optimization formulations, the PSO algorithm, and MPS, applied to two chemical process examples, are shown through numerical simulations.