Support Vector Machine and TOPSIS Based Driving Style Classification Judgment
杨童①③YANG Tong;王猛②WANG Meng;彭国栋① PENG Guo-dong;陈瑶①CHEN Yao
①云南省交通科学研究院有限公司,昆明650011;
②昆明理工大学交通工程学院,昆明650500;
③云南省数字交通重点实验室,昆明650200
①Yunnan 'Transportation Science Research Institute Co.,Ltd.,Kunming 650011 , China ;
②Faculty of Transportation Engineering,Kunming University of Science and Technology, Kunming 650500,China;
③Yunnan Provincial Key Laboratory of Digital Transportation,Kunming 650200, China
摘要;不同驾驶风格对驾驶安全产生重要的影响,针对各类驾驶风格做出安全性综合评价,通过实车试验采集50名驾驶人的行车数据,结合专家对被试者驾驶风格的评判并利用支持向量机算法对驾驶风格进行分类,构建TOPSIS评价法对各类驾驶风格进行综合安全评估并提出建议。结果表明,支持向量机对驾驶风格分类结果准确率高达80.5%,保守型驾驶风格的安全性最高.
Abstract: Different driving styles have an important impact on driving safety. A comprehensive evaluation of safety is made for eachtype of driving style. The driving data of 50 drivers were collected through real est, and the 'TOPSIS evaluation method was constructed tomake a comprehensive safety evaluation of various driving styles by combining expert judgments on the driving styles of the subjects andusing SVM algorithms to classify the driving styles and give advice. The results showed that support vector machine classification of drivingstyle results with an accuracy of 80.5%, conservative driving style being the safest.
关键词:驾驶风格;支持向量机;专家评判法; TOPSIS评价模型
Key words: driving style; SVM; expert judgment method; TOPSIS
|