用神经网络评价船舶碰撞危险的研究

Collision Risk Assessment Using Neural Network

笔者将人工神经网络的ＢＰ学习算法用于船舶碰撞危险评价的研究，试图用神经网络的ＢＰ算法代替模糊推理系统。结果表明，神经网络可以用于碰撞危险的评价，也适用于避碰决策系统的研究。

Authors use back propagation learning algorithm of neural network to make assessment of ship collision risk. Compared with the results from the fuzzy inference system, the neural network method is much faster, and it can be used to replace the fuzzy inference system in collision avoidance expert system.