Data‐driven prescriptive maintenance toward fault‐tolerant multiparametric control

Abstract

Prescriptive maintenance can improve system effectiveness and system safety via integrated production and maintenance optimization. However due to system disruptions there is potential for abnormal operations and an undesirable increased occurrence of process safety incidents. This research provides a multiparametric‐based framework for safety‐aware, maintenance‐aware, and disruption‐aware process control. It leverages ensemble classification via machine learning classifiers for fault detection, mixed‐integer nonlinear programming for integrated safety‐aware production and maintenance scheduling, as well as hybrid multiparametric model predictive control for fault‐tolerant setpoint tracking. The results show that the ensemble classifier outperforms the individual classifiers in terms of fault detection accuracy, sensitivity, and specificity. Furthermore, it is seen that the developed controllers are able to reconfigure the control actions based on process disruption information. The framework is illustrated with a chemical complex system, and a cooling water system. The approach can be used to help improve the safety and productivity of industrial processes.

Design of Big Data Task Scheduling Optimization Algorithm Based on Improved Deep Q-Network

“…V refers to the current task. The execution time is represented by the ECT matrix of VP  , which is the time required to allocate each task to different nodes at the current moment [12]. In scheduling problems, the most common definitions are the earliest start time, earliest completion time, and maximum completion time.…”

Multi-agent deep reinforcement learning based Predictive Maintenance on parallel machines

“…Math Prog. ML [27] 2022 GA [28] 2021 BH, VNS [29] 2021 GA, MC [30] 2021 GA [31] 2021 GA [32] 2021 GrA [33] 2021 GA [34] 2021 CEA [35] 2021 GrA [36] 2021 GA [37] 2021 BD [38] 2021 RHA [12] 2021 MP NN [39] 2021 RL [40] 2021 RL [41] 2020 GA [42] 2020 MVO [43] 2020 TS [44] 2020 NSGA-II [45] 2020 ANSGA-III [46] 2020 MA MIP [47] 2020 GA MIP [48] 2020 BOMP [49] 2020 BB [50] 2020 RL [51] 2020 RL [52] 2020 NN [53] 2019 GA [54] 2019 OTA [55] 2019 AC [56] 2019 NSGA-II [57] 2019 BIP [58] 2019 MILP [59] 2019 MINLP [60] 2019 RL [61] 2019 RL [62] 2019 RL [10] 2018 GA [63] 2018 SA, GA [64] 2018 SA [65] 2017 NSGA-II [66] 2017 MA [67] 2017 MILP [68] 2017 RL [69] 2016 NSGA-II [70] 2016 LPT MILP…”

Prescriptive maintenance for complex products with digital twin considering production planning and resource constraints

“…Elbasheer et al [10] suggested to incorporate RxM into production planning, and developed an intelligent decision support agent based on reinforcement learning. To cope with abnormal operations caused by system disruptions, Gordon and Pistikopoulos [11] proposed a multiparametric-based framework, which adopts ensemble classification for fault detection and mixed-integer nonlinear programming for safety-aware production and maintenance scheduling. However, for real production environments, production planning along with equipment status are often accompanied by a high degree of uncertainty and dynamics, which is a considerable challenge for maintenance program development.…”