The invention discloses a road traffic accident risk prediction deep learning algorithm, which comprises the following steps: S1, feature extraction: processing a multi-source traffic data set by using a convolutional neural network algorithm, a long-short-term memory neural network algorithm and a convolutional long-short-term memory neural network algorithm, and performing feature extraction; classifying the high-correlation multi-source features by using a random forest network algorithm to analyze features inducing traffic accidents, and obtaining a deep neural network of traffic accident occurrence risks; s2, establishing a random forest neural network prediction model, and adopting the random forest neural network prediction model containing a plurality of decision trees to perform classification prediction on the traffic accident risk; and S3, simulation result and analysis: establishing a simulation environment, performing a simulation experiment by using simulation software, experimental data and traffic accident data of a certain region in recent years and using a neural network library, and analyzing a result. According to the invention, the accuracy of road traffic accident risk prediction can be effectively improved.
本发明公开了道路交通事故风险预测深度学习算法,包括S1、特征提取,利用卷积神经网络算法、长短期记忆神经网络算法、卷积长短期记忆神经网络算法处理多源交通数据集,进行特征提取,使用随机森林网络算法对高相关性多源特征进行分类,以分析诱发交通事故的特征,并得到交通事故发生风险性的深度神经网络;S2、建立随机森林神经网络预测模型,采用含有多个决策树的随机森林神经网络预测模型对交通事故风险做分类预测;S3、仿真结果与分析,搭建仿真环境,利用仿真软件、实验数据,利用某地区近几年交通事故数据,使用神经网络库,进行仿真实验,并分析结果。本发明能够有效提高道路交通事故风险预测的精度。
Road traffic accident risk prediction deep learning algorithm
道路交通事故风险预测深度学习算法
2022-04-12
Patent
Electronic Resource
Chinese
A Deep Learning Approach to the Citywide Traffic Accident Risk Prediction
ArXiv | 2017
|Traffic Accident Risk Prediction Using Machine Learning
IEEE | 2022
|