基于SVM的弯道行车安全辨识  

Safety Identification of Curved Road Based on SVM

作　　者：胡婧晖 宇仁德[1] 王道意 崔淑艳 朱燕华 闫兴奎 HU Jinghui;YU Rende;WANG Daoyi;CUI Shuyan;ZHU Yanhua;YAN Xingkui(School of Transportation and Vehicle Engineering,Shandong University of Technology,Zibo Shandong 255049)

机构地区：[1]山东理工大学交通与车辆工程学院,山东淄博255049

出　　处：《河南科技》2021年第20期9-12,共4页Henan Science and Technology

摘　　要：车辆运行安全状态辨识能够很好地提前规避风险,在弯道这样事故多发的路形上显得尤为重要。通过ADAMS软件仿真车速、弯道半径、超高对车辆过弯行驶状态的影响并进行预处理,形成辨识模型的训练集,再分别用网格搜索法、遗传算法、粒子群算法对支持向量机(Support Vector Machines,SVM)模型进行优化,建立安全状态辨识模型。试验结果表明,优化后的SVM模型可以有效辨识出弯道行驶安全状态,识别率超过90%。The identification of the safe state of vehicle operation can avoid risks well in advance,which is particularly important in the shape of roads with frequent accidents such as curves.The ADAMS software is used to simulate the influence of vehicle speed,curve radius,and superelevation on the vehicle’s cornering driving state and carry out preprocessing,the training set of the identification model is formed,and the SVM model is optimized by the grid search method,genetic algorithm,and particle swarm algorithm to establish a safety state identification model.Experimental results show that the optimized SVM model can effectively identify the safe state of cornering,with a recognition rate of over 90%.

关 键 词：弯道行车安全性 支持向量机 辨识 

分 类 号：U463.6[机械工程—车辆工程]