基于机器学习算法的船舶碰撞事故等级预测研究  

Research on Prediction of Ship Collision Level Based on Machine Learning Algorithm

作　　者：杨欣怡 李为为 YANG Xinyi;LI Weiwei(College of Traffic and Transportation, Chongqing Jiaotong University, Chongqing 400074, China)

机构地区：[1]重庆交通大学,交通运输学院,重庆400074

出　　处：《微型电脑应用》2022年第2期116-119,共4页Microcomputer Applications

摘　　要：为了更好地了解水上事故影响因素与事故等级之间的联系,精准地预测水上船舶碰撞事故的严重程度,从人、船、环境、管理4个方面考虑,结合2000年到2020年之间的54个典型长江船舶碰撞事故案例,量化事故影响因素,提出了一种基于RF-BP神经网络的预测模型。结果表明,该方法预测结果与实际事故结果之间误差较小,与传统的BP神经网络模型相比训练时间更短,且精度未降低。可见该方法具有精度高、训练速度快的特点,在多因素事故等级预测中具有一定的可行性与实用性。In order to better understand the connection between the influencing factors of water accident and accident level,and predict the severity of the water of ship collision accident accurately,a forecasting model based on RF-BP neural network model is proposed by considering four aspects of human,ship,environment and management,and quantifying the accident of influencing factors in 54 typical ship collision accidents in Yangtze river between 2000 to 2020.The results show that error between the prediction result and the actual accident result is little,and the model is compared with the traditional BP neural network model,the results are that the training time is shorter,and the accuracy is not reduced.It is thus clear that this method has the characteristics of high precision and fast training speed,and has certain feasibility and practicability in multi-factor accident level prediction.

关 键 词：水上交通事故 数据挖掘 事故预测 BP神经网络 随机森林 

分 类 号：U698.6[交通运输工程—港口、海岸及近海工程] TP181[交通运输工程—船舶与海洋工程]