Self-adaptive optimized maintenance of offshore wind turbines by intelligent Petri nets

Grouping Maintenance Policy for Improving Reliability of Wind Turbine Systems Considering Variable Cost

“…Saleh et al [13] developed the self-adaptive maintenance policy for offshore wind turbines by intelligent Petri nets, which avoids dispensable maintenance behaviors and reduces the operation and maintenance costs related to downtime. Ade Irawan et al [14] used service operation vessels and safe transfer boats to make maintenance routing in offshore wind farms.…”

Machine Learning Solutions for Offshore Wind Farms: A Review of Applications and Impacts

“…For the classification analysis, the authors used linear discriminate analysis, SVM, KNN, random forest, naive Bayes, and decision tree, where the first three could deliver the most accurate results. In another study, Saleh et al combined advanced Petri net modeling with reinforcement learning, resulting in a versatile methodology applicable to optimizing various Petri net models [132]. The method, called intelligent Petri net, fuses reinforcement learning techniques with Petri net principles.…”

A sequential decision problem formulation and deep reinforcement learning solution of the optimization of O&M of cyber-physical energy systems (CPESs) for reliable and safe power production and supply

“…Recently, many researches have focused on O&M decision making: to name few, in [18] an artificial neural network is proposed to estimate the maintenance cost and, then used within a multi-agent Deep Reinforcement Learning (DRL) model to optimize decisions on large-scale systems; in [19] a Petri Net is applied to optimize offshore wind turbines O&M; in [20], a Bayesian Network maximizes a system supply capacity and gas supply reliability within a DRL scheme for maintenance planning. In all cases, however the fluctuations of the energy production and demands, their uncertainty, especially under increasing scenarios of penetration of RES specific production plans, have been overlooked.…”

“…In a SDP, the goodness of the selected O&M action does not depend exclusively on the actual decision, but, rather, on the whole sequence of future decisions. To solve the SDP for the optimal O&M sequence of actions, we rely, as in [18][19][20], on Deep Reinforcement Learning (DRL), which is an extension of Reinforcement Learning (RL) and provides feasible application to complex systems [21,22]. RL has been applied to complex decision-making problems in many fields, including energy-related ones [23][24][25][26][27][28][29][30][31][32].…”