基于SSA-SVM的道路事故预测模型
Road Accident Prediction Model Based on SSA-SVM
杨朗1
作者信息
- 1. 湖南陆德工程咨询有限公司,湖南 衡阳 421200
- 折叠
摘要
将道路事故特征因素和环境因素作为自变量,事故严重程度作为因变量,采用SVM构建道路事故预测模型,并使用SSA优化模型中的最优参数组合,最终建立基于SSA优化的SVM道路事故预测模型,模型预测准确率优于广义线性预测模型和BP神经网络预测模型.经过模型致因分析,得到三项对道路事故严重程度具有显著影响的自变量,分别为机非隔离形式、监控设施和最近交叉口交通组织形式,可作为道路交通安全管控的核心控制参数.
Abstract
With road accident characteristic factors and environmental factors as independent variables and accident severity as depend-ent variables,SVM was used to build a road accident prediction model,and the optimal parameter combination in the SSA optimization model was used to establish a road accident prediction model based on SSA optimization.The prediction accuracy of the model was bet-ter than that of generalized linear prediction model and BP neural network prediction model.Through the model causation analysis,three independent variables that have a significant impact on the severity of road accidents are obtained,namely,the non-isolation form,monitoring facilities and the traffic organization form of the nearest intersection,which can be used as the core control parameters of road traffic safety management and control.
关键词
道路事故预测/SVM模型/SSA优化/致因分析Key words
road accident prediction/SVM model/SSA optimization/causation analysis引用本文复制引用
出版年
2024
NETL
NSTL
万方数据