基于二次Lasso回归的纵向驾驶员模型研究
Research on Longitudinal Driver Model Based on Quadratic-Lasso Regression
投稿时间:2021-03-20  修订日期:2021-09-07
DOI:
中文关键词:  纵向驾驶员模型  二次回归  Lasso回归  工况跟随
英文关键词:Longitudinal driver model  Quadratic regression  Lasso regression  Driving cycle following
基金项目:国家重点研发计划项目(2018YFB0105900), National Key R D Program of China (2018YFB0105900)
作者单位邮编
刘通 吉林大学 汽车仿真与控制国家重点实验室 130025
曾小华 吉林大学 汽车仿真与控制国家重点实验室 130025
李泰祥 中车青岛四方车辆研究所有限公司 
宋大凤 吉林大学 汽车仿真与控制国家重点实验室 
庄晓 一汽解放青岛汽车有限公司 
摘要点击次数: 14
全文下载次数: 0
中文摘要:
      为了更好地模拟实际驾驶员行为提出了基于二次回归和Lasso回归方法的纵向驾驶员回归模型。首先,通过采集纵向驾驶行为数据,提取可能影响驾驶行为的状态参数,进而建立二次回归驾驶员模型;然后,面向多参数回归模型中的多重共线问题,采用Lasso回归方法进行状态参数筛选;最后,结合筛选数据建立二次回归驾驶员模型。为了验证模型的有效性,与PI驾驶员模型和一次Lasso回归驾驶员模型进行了仿真对比。仿真结果表明,相比于其他两种模型,所建立的驾驶员模型具备良好的工况跟随效果,同时能较好地反映实际驾驶行为特征。
英文摘要:
      The driver model plays an important role in vehicle control strategy design. To better simulate real driver behavior, a longitudinal driver regression model based on quadratic regression and Lasso regression method is proposed. Firstly, by collecting longitudinal driving behavior data, the state parameters able to affect driving behavior are extracted before the quadratic regression driver model is established. Then, the Lasso regression method is applied to screen the state parameters for the multicollinearity problem in the multi-parameter regression model. And finally, a quadratic regression driver model is developed based on the screened data. In order to verify the effectiveness of the model, a simulation comparison was made between the proposed model and PI driver and Lasso regression driver models. The simulation results showed the driver model developed here to not only have a better effect on driving cycle traction compared to the other two models, but also better reflect the characteristics of actual driving behavior.
  查看/发表评论  下载PDF阅读器
关闭