Assessment of the vehicle handling especially with respect to its lateral dynamics is an important aspect of the overall vehicle design and development process. However, an increasing rate of vehicle update cycles, a massively growing number of variants and a required high-quality comfort and/or driving reward render the evaluation process a challenging task. Consequently, virtual methods support the overall development process and increase time and cost efficiency significantly. The so-called model-based (objective) handling methodology aims to extract certain vehicle and/or driver model parameters from measurement data. These can then be used to simulate standard handling maneuvers, rather than performing them on a test track. State-of-the-art parameter identification mechanisms are commonly performed offline and require extensive instrumentation of the test vehicle. This chapter presents a novel approach exploiting observer-based parameter identification techniques. It introduces the advantages of online capability, time-efficient experiment execution and reduction of sensing devices due to estimation of specific system states. The joint estimation of states and parameters is formulated as an unknown input recovery problem. Using sliding mode mechanisms allows formulation of state observers that are invariant with respect to certain classes of perturbations. Furthermore, the attractiveness is increased considering the property of finite time convergence. Herein, higher-order sliding mode concepts are used for the task of parameter identification providing robustness, finite convergence time and stability even for non-persistently excited systems. Evaluation of the concepts is performed in a twofold way: (a)An industrial vehicle dynamics simulation tool provides data for the observation concepts. The resulting parameter estimates are integrated into the offline simulation of standard handling maneuvers, e.g., step input steering. Comparing these results with the reference data allows to draw conclusions on the expected accuracy of the method. (b) The selected concepts are evaluated on experimental facilities, i.e., standard vehicles and an electric power steering test bench.

Chapter Contents:

• 7.1 Introduction
• 7.1.1 Modeling of selected vehicle dynamics
• 7.1.1.1 Lateral vehicle motion
• 7.1.2 Chassis roll motion
• 7.1.3 Model parameters and sensor setup
• 7.2 State-of-the-art review
• 7.3 Contributions
• 7.3.1 Robust lateral dynamics parameter estimator
• 7.3.1.1 Lateral tire force observer
• 7.3.1.2 Lateral cornering stiffness observer
• 7.3.2 Robust roll dynamics parameter estimator
• 7.3.2.1 Robust exact differentiator (RED)
• 7.3.2.2 Robust state estimation (RSE)
• 7.3.2.3 Adaptive robust state estimation (ARSE)
• 7.3.2.4 Parameter identification (PE)
• 7.3.2.5 Adaptive roll angle observation (ARAO)
• 7.3.3 Simulation-based concept evaluation
• 7.3.3.1 Parameter identification results
• 7.3.3.2 Handling evaluation results
• 7.3.4 Experimental results
• 7.4 Conclusions
• Acknowledgment
• References

Inspec keywords: vehicle dynamics; variable structure systems; steering systems; parameter estimation

Other keywords: finite time convergence; vehicle dynamics assessment; time-efficient experiment execution; online capability; finite convergence time; electric power steering; input recovery problem; standard handling maneuvers; sensing devices reduction; higher-order sliding mode concepts; observer-based parameter identification; sliding mode mechanisms

Subjects: Transportation system control; Vehicle mechanics; Multivariable control systems