| 基于交互多模型的车辆质量与道路坡度估计（双语出版） |
| |
| 引用本文： | 赵健,李至轩,朱冰,李雅欣,孙玉泽.基于交互多模型的车辆质量与道路坡度估计（双语出版）[J].中国公路学报,2019,32(12):58-65. |
| |
| 作者姓名： | 赵健 李至轩 朱冰 李雅欣 孙玉泽 |
| |
| 作者单位： | 吉林大学 汽车仿真与控制国家重点实验室, 吉林 长春 130022 |
| |
| 基金项目： | 国家重点研发计划项目（2018YFB0105100）；国家自然科学基金项目（51575225，51775235）；吉林大学高层次科技创新团队项目（2017TD-20） |
| |
| 摘 要： | 车辆结构参数和道路环境信息的实时准确获取是提高智能汽车运动控制性能的重要因素之一，而车辆质量与道路坡度信息是多种汽车控制系统的必要信息，因此质量与坡度在线估计的研究一直受到关注。针对车辆质量与道路坡度的联合估计问题，提出了一种基于交互多模型的质量与坡度融合估计方法。首先，设定了适宜进行质量精确估计的工况条件，据此提出了基于模糊规则的质量估计置信度因子计算算法，进而设计了基于置信度因子的递推最小二乘车辆质量估计算法，以实现质量的在线估计。然后，以车辆纵向动力学模型为基础，建立了运动学和动力学2种坡度估计模型，并设计了基于运动学模型的线性卡尔曼滤波坡度观测器，基于电子稳定性程序ESP的纵向加速度信息实现坡度估计，设计了基于动力学模型的无迹卡尔曼滤波坡度观测器，基于ESP和发动机管理系统EMS的力信息实现坡度估计。运动学模型未考虑车辆姿态信息，坡度估算结果与实际值有偏差；动力学模型对模型精度要求高，算法稳定性差，为充分发挥2种方法优势实现坡度的精确估计，采用交互多模型算法实现了2种坡度估计方法的加权融合。最后，对所设计的算法进行了实车试验验证。结果表明：所设计的质量与坡度估算算法具有较好的实时性和准确性，适合智能汽车运动控制的应用需求。
|
| 关 键 词： | 汽车工程 质量与坡度估计 质量估计置信度因子 交互多模型融合 智能估计 实车试验 |
| 收稿时间： | 2019-04-05 |
Vehicle Mass and Road Slope Estimation Based on Interactive Multi-model（in English） |
| |
| ZHAO Jian,LI Zhi-xuan,ZHU Bing,LI Ya-xin,SUN Yu-ze.Vehicle Mass and Road Slope Estimation Based on Interactive Multi-model（in English）[J].China Journal of Highway and Transport,2019,32(12):58-65. |
| |
| Authors: | ZHAO Jian LI Zhi-xuan ZHU Bing LI Ya-xin SUN Yu-ze |
| |
| Affiliation: | State Key Laboratory of Automotive Simulation and Control, Jilin University, Changchun 130022, Jilin, China |
| |
| Abstract: | Real-time and accurate acquisition of vehicle structural parameters and road environment information is important to improve the performance of intelligent vehicle motion control. Vehicle mass and road slope are necessary information for various vehicle control systems. Therefore, research on the online estimation of mass and slope has always been concerned. Aiming at the joint estimation of vehicle mass and road slope, an estimation method based on interactive multi-model is proposed in this paper. Firstly, the working conditions suitable for accurate mass estimation were established. Under these conditions, an algorithm for calculating the confidence factor of mass estimation based on fuzzy rules was proposed. Then, a recursive least squares vehicle mass estimation algorithm based on a confidence factor was designed to realize online mass estimation. Based on the vehicle longitudinal dynamics model, two kinds of slope estimation models, kinematics, and dynamics were established. The linear Kalman filter slope observer based on the kinematics model was designed to realize slope estimation based on the longitudinal acceleration information of ESP (Electronic Stability Program). The unscented Kalman filter slope observer based on the dynamic model was designed to realize slope estimation according to the force information of ESP and EMS (Engine Manage System). The kinematics model did not consider the vehicle attitude information, and the slope estimation results deviated from the actual values. The dynamic model required high precision and had poor stability. In order to make full use of the advantages of the two methods and achieve accurate slope estimation, the weighted fusion of the two methods was realized using the interactive multiple model algorithm. Finally, the algorithm was verified by vehicle test. The results show that the mass and slope estimation algorithm has good real-time performance and accuracy, and meets the application requirements for intelligent vehicle motion control. |
| |
| Keywords: | automotive engineering mass and slope estimation confidence factor of mass estimation interactive multi-model fusion intelligent estimation real vehicle test |
|
| 点击此处可从《中国公路学报》浏览原始摘要信息 |
|
点击此处可从《中国公路学报》下载全文 |
|
