Answers are grounded in this paper and cite their sources.
This study presents a novel approach for estimating the dynamic tire normal force on all four wheels using a gated recurrent unit (GRU) model. Advanced driver assistance systems are among the key technologies employed for enhancing driving stability and vehicle safety, and the tire normal force is a crucial factor in their design and analysis. However, direct measurement of this force in passenger vehicles is extremely challenging due to high costs, technical limitations, and the burden of maintenance. Furthermore, conventional Kalman filter (KF)–based estimation approaches suffer from performance degradation caused by model uncertainties and variations in identified parameters under diverse driving conditions. The proposed GRU model can capture the dynamic tire normal force using only time-series data acquired from conventional onboard vehicle sensors, without requiring any explicit physical model. As an initial application, a preliminary study on hydroplaning classification from estimated dynamic tire normal force responses is also presented; hydroplaning is a physical phenomenon which leads to loss of tire-road contact and a reduction in tire normal force. In this study, the estimated tire normal force is utilized as a signature signal for hydroplaning classification based on one-dimensional convolutional neural networks. The estimation performance of the proposed GRU model is comprehensively validated through comparisons with a KF–based estimator, CarSim®vehicle simulations, and in-vehicle experiments with a smart tire sensor, and the hydroplaning classification is further verified using an embedded edge-computing module.
No takes yet. Share an insight, caveat, or question.
Lee et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: