A cascaded scheme for high-performance estimation of vehicle states

Abstract

Vehicle active safety control is bonded tightly with the accurate acquirement of vehicle states. This paper presents a cascaded scheme to realize high-performance estimation of vehicle states. To achieve the estimation with good performance, an adaptive sliding mode observer is designed for determining four longitudinal tire forces independently, and the Kalman filter is used for alleviating the inherent chattering effect. On this basis, lateral tire forces are calculated via a simplified formula based on the Dugoff tire model. Lastly, utilizing the obtained tire force information, the key states of vehicle motion are estimated through the smooth variable structure filter. Numerical experiments are conducted to testify the effectiveness of the presented estimation scheme. The results of performance comparison in different case studies show that the chattering effect can be suppressed to a great extent, and the accuracy, robustness and real-time performance to modeling uncertainty and unexpected measurements can be effectively guaranteed for vehicle state estimation by means of the proposed scheme.

Vehicle State Estimation Based on Multidimensional Information Fusion

“…Therefore, the maximum number of edges Nmax acquired in a single period should be limited first, as shown in Eq. (2).…”

“…Vehicle driving state estimation not only serves dynamics control [1], but also provides important help for intelligent driving decision [2]. Its accuracy and real-time performance directly affect the actual effect of system functions, and even determine the stability and safety of the vehicle [3,4].…”