Robustness aspects in the design of a vehicle state estimator

In a vehicle, many parameters, such as the sprung mass, are not fixed. Rather, they change in a specified range. Vehicle control has to be able to cope with these potentially large parameter variations in the system. This paper deals with an algorithm, used to control the body slip angle. An estimator, part of the control, is used to obtain non-measurable signals. Whereas the vehicle parameters are uncertain, the estimator comprises of a vehicle model with fixed parameters. The objective is to find these nominal parameters, that guarantee optimal performance and stability in the complete parameter space of the vehicle. This requirement is defined in mathematics as an optimisation task. Even considering only the simple linear case, the problem very soon reaches such a complexity that it demands computational support. Nonlinear influences such as the tyre characteristics and weighting functions make analytical solutions either impossible or too restrictive for the design engineer.