Risk assessment of refinery unit maintenance based on fuzzy second generation curvelet neural network
Bin Zhao, Hao Chen, Diankui Gao, Lizhi Xu
Abstract
To ensure safety of maintenance of refinery unit and improve reliability of refinery unit, the risk assessment model of refinery unit maintenance is established based on fuzzy second generation curvelet neural network optimized by improved firefly algorithm. Firstly, the fuzzy second generation curvelet neural network is constructed based on second generation curvelet transform, artificial neural network, and fuzzy theory. Secondly, the training algorithm of fuzzy second curvelet neural network is designed based on improved firefly algorithm, the traditional firefly algorithm is improved based on chaotic mapping. And the corresponding algorithm procedure is designed. Finally thirty-five gasoline hydrodesulphurization units of petrochemical companies are used as research objects to carry out risk assessment of maintenance based on the proposed method and other two existing methods, results show that the proposed fuzzy second curvelet neural network has better computing precision and efficiency, which can correctly and efficiently evaluate the risk level of maintenance of refinery unit.