摘要
道路交通事故预测作为道路主动安全管理的重要组成部分,在降低事故发生概率、帮助管理者制定安全决策等方面起着重要作用。随着数据需求的不断增加,传统方法已无法满足大数据的需求,机器学习和人工智能算法在动态、实时和复杂情况下的道路交通事故预测领域显示出强大的潜力。文章从数据获取和特征变量选择开始介绍,详细叙述了基于机器学习的神经网络及与深度学习结合后该方法在国内外的相关研究,分析了使用神经网络相关方法在建模时会面临的优缺点,最后对基于神经网络的交通实时事故预测方法进行了总结及展望,给出未来的发展趋势。
As an important part of active road safety management,road traffic accident prediction plays an important role in reducing the probability of accidents and helping managers make safety decisions.With the increasing demand for data,traditional methods can no longer meet the needs of big data,and machine learning and artificial intelligence algorithms have shown strong potential in the field of road traffic accident prediction in dynamic,real-time and complex situations.This paper introduces the data acquisition and the selection of characteristic variables,describes in detail the relevant research of the neural network based on machine learning and the combination of the method with deep learning at home and abroad,analyzes the advantages and disadvantages of using the neural network correlation method in modeling,and finally summarizes and looks forward to the realtime traffic accident prediction method based on neural network,and gives the future development trend.
作者
炎天策
YAN Tiance(School of Automobile,Chang'an University,Xi'an 710064,China)
出处
《汽车实用技术》
2024年第5期176-181,共6页
Automobile Applied Technology
关键词
交通工程
实时事故预测
神经网络
深度学习
Traffic engineering
Real-time accident prediction
Neural network
Deep learning