| 基于BP神经网络的智能车辆换道决策模型研究 |
| |
| 作者姓名： | 陈力 殷时蓉 罗天洪 郑讯佳 |
| |
| 作者单位： | 重庆交通大学机电与车辆工程学院;重庆文理学院智能制造工程学院;汽车噪声振动和安全技术国家重点实验室;中国汽车工程研究院股份有限公司 |
| |
| 基金项目： | 重庆市教委科技项目(KJQN)。 |
| |
| 摘 要： | 为了正确刻画智能网联环境下的车辆换道行为,提出基于BP神经网络的车辆换道决策模型.分析了交通流中车辆换道行为,以HighD自然驾驶数据集为数据来源,筛选出1 900组车辆换道和未换道信息作为模型的训练与验证,利用高斯滤波方法拟合目标车辆换道轨迹和横向位移轨迹,选择影响车辆换道决策的7个参数作为模型输入,建立BP神经网络...
|
| 关 键 词： | 智能交通 车辆换道 BP神经网络 换道决策模型 智能汽车 |
Research on Lane-Changing Decision Model for Intelligent Vehicles Based on BP Neural Network |
| |
| Authors: | CHEN Li YIN Shirong LUO Tianhong ZHENG Xunjia |
| |
| Institution: | (School of Mechatronics and Vehicle Engineering,Chongqing Jiaotong University,Chongqing 400074,China;School of Intelligent Manufacturing Engineering,Chongqing University of Arts and Science,Chongqing 402160,China;State Key Laboratory of Automotive NVH and Safety Technology,Chongqing 401122,China;China Automotive Engineering Research Institute Co.,Ltd.,Chongqing 401122,China) |
| |
| Abstract: | In order to correctly describe the lane-changing behavior of vehicles in the intelligent networked environment, a vehicle lane-changing decision model based on BP neural network is proposed. Initially, the lane-changing behavior in traffic flow was analyzed, and then the HighD natural driving dataset was used to select 1 900 samples with or without lane-changing behavior for model training and testing. The Gaussian filtering was applied to fit the lane-changing trajectories and the lateral displacement trajectories. 7 parameters that affect the lane-changing decisions were chosen as the model input. Finally, the BP neural
network based lane changing decision model was trained and tested. The results show that the decision accuracy of the proposed BP neural network model reaches 96.2%, which can effectively identify the lane changing behavior of vehicles. |
| |
| Keywords: | intelligent transportation vehicle lane changing BP neural network lane changing decision model intelligent car |
| 本文献已被 维普 等数据库收录！ |
| 点击此处可从《》浏览原始摘要信息 |
| 点击此处可从《》下载免费的PDF全文 |
|
