Model‐predictive safety optimal actions to detect and handle process operation hazards

Abstract

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. K E Y W O R D Schemical processes, model-predictive safety, predictive alarm, process constraints, process safety, receding horizon

State‐of‐the‐Art Review on the Applications of Nonlinear and Artificial Intelligence‐Based Controllers in Petrochemical Processes

“…Petrochemical processes are characterized by their intricate and dynamic behavior, which arises from the inherent variability of the process inputs and the intricate interactions between the different process variables. Understanding and predicting this behavior is essential for ensuring that the process operates safely, efficiently, and within the desired operating range [40,41]. To address these process behaviors, petrochemical processes use a variety of control techniques, such as feedback control, feedforward control, and advanced control algorithms, to keep the process variables at desired levels that fall within acceptable ranges [42][43][44].…”

Ti3C2MXene–polymer nanocomposites and their applications

“…In every technology, process safety is of prime importance. Despite advances in process safety and the introduction of increasing tighter safety regulations, more than 50 serious incidents happened in the U.S. over the past ten years 244 . Product safety is also of great importance, as the users of a product and the environment should not be harmed by the product.…”

Adversarially robust Bayesian optimization for efficient auto‐tuning of generic control structures under uncertainty

“…We demonstrate the value of the proposed ARBO method on two case studies; an illustrative problem to highlight the key steps and advantages of ARBO and a challenging auto‐tuning problem in which a highly nonlinear bioreactor with several unknown parameters is controlled using nonlinear MPC (NMPC) with multiple constraint backoffs that must be tuned. It is important to note that, although we focus on nominal MPC, the ARBO can robustly auto‐tune any controller structure including those that incorporate recently proposed safety schemes such as the model‐predictive safety (MPS) method 18,19 …”

“…It is important to note that, although we focus on nominal MPC, the ARBO can robustly autotune any controller structure including those that incorporate recently proposed safety schemes such as the model-predictive safety (MPS) method. 18,19…”