The Synthetic Assessment Modeling of Ships' Oil Spill Risk based on Fuzzy Neural Network

X. Cao and S. Fan (PRC)


risk management, risk assessment, fuzzy neural network, and oil spill risk


The paper firstly emphasizes the importance of synthetic risk assessment of ship oil spill risk which would be useful for the prevention and management of oil spill, and be beneficial to provide effective help and policy decision. By analysing the current research, neglecting the environmental hazard assessment and management deficiencies is the congenital defect. Ships′ oil spill accidents are multi-dimensional systems which constitute many factors such as time, space, natural factors, shipping factors, channel factors, transport factors, as well as crew factors. The characteristics of system are rather particular, because ship oil spill accidents frequently occur having the protruding property of the randomness and uncertainty. So traditional comprehensive evaluation method are not very good solution to the complex and non-linear systems, but coupling theory merges with neural network techniques and fuzzy mathematics can across these restrictions. The paper builds risk assessment model of ships′ oil spill based on fuzzy neural network which relates two emphasis, namely, index system and coupling assessment model. The relevant maritime data is input into the network ,and by training, the network parameters and structure are optimized, achieving effective safety evaluation.

Important Links:

Go Back