Scholarlink

<p>The vehicle mass is a crucial parameter for analyzing operational conditions and maintenance needs from multi-source heterogeneous vehicle data. For the mass estimation of light trucks, a vehicle longitudinal dynamics model that considers the coupling of sensor nonlinear noise has been established. Furthermore, this paper optimizes the Unscented Kalman Smoother (UKS) algorithm by considering dynamic time steps and backward parameter propagation. The estimation method uses unstructured basic vehicle parameters and structured onboard Controller Area Network (CAN) data as inputs to dynamically identify vehicle status and output joint estimates of truck mass and slope. Using real road sections in Nanchang as examples, we conducted multiple rounds of testing on this estimation method. Results show that the estimation accuracy of this method can exceed 95%, with convergence speeds reaching seconds-level granularity. In addition, the estimation technique achieves greater speed and accuracy in scenarios with unloaded and fully loaded; however, in cases of partial loading, there is a noticeable increase in variability of the recognition outcomes, along with a comparatively extended settling time.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Joint Estimation of Truck Mass and Road Slope Based on the Unscented Kalman Smoother

  • 期刊论文

  • 发表日期：

    2025年11月19日

查看全文
  • Ning Wang,
  • Xiufeng Li,
  • Liaodong Nie,
  • Kuan Kei Hoi

摘要

The vehicle mass is a crucial parameter for analyzing operational conditions and maintenance needs from multi-source heterogeneous vehicle data. For the mass estimation of light trucks, a vehicle longitudinal dynamics model that considers the coupling of sensor nonlinear noise has been established. Furthermore, this paper optimizes the Unscented Kalman Smoother (UKS) algorithm by considering dynamic time steps and backward parameter propagation. The estimation method uses unstructured basic vehicle parameters and structured onboard Controller Area Network (CAN) data as inputs to dynamically identify vehicle status and output joint estimates of truck mass and slope. Using real road sections in Nanchang as examples, we conducted multiple rounds of testing on this estimation method. Results show that the estimation accuracy of this method can exceed 95%, with convergence speeds reaching seconds-level granularity. In addition, the estimation technique achieves greater speed and accuracy in scenarios with unloaded and fully loaded; however, in cases of partial loading, there is a noticeable increase in variability of the recognition outcomes, along with a comparatively extended settling time.