Comparison of Model-Based Techniques for Vehicle Sideslip Angle Estimation

In the automotive industry, on-board vehicle estimation of the side-slip angle plays a fundamental role in applications related to autonomous driving and within the design of active safety systems for controlling vehicle lateral dynamics. To ensure the availability of the real-time monitoring of this quantity, which can only be measured through sophisticated and expensive systems, various virtual sensing techniques have been developed in the literature. In this scenario, it is essential to find a balance between estimation robustness and computational burden to ensure the algorithm to be employable in embedded devices for on-board applications. The current paper proposes analytical and graphical analysis aimed at examining the estimation accuracy of various physical model-based methodologies, differing in terms of numerical complexity and computational cost, but employing equal tire and vehicle modelling formulation. The numerical results have been compared with accurate side-slip angle measurements, acquired during on-track test sessions, which also included the low-end sensors’ signals employed into the proposed estimators.