Page - 36 - in Maximum Tire-Road Friction Coefficient Estimation

2 Estimation of the friction potential Table 2.3.: Criteria and weighting factors used and described in detail in Lex et al., [LEH11] Criterion Weighting Factor Friction potential (in contrast to demanded coeff. of friction) 8 Online capability of algorithm 7 No additional sensors needed 6 Active intervention not necessary for robust estimate 5 Applicability in ADAS 4 Availability of estimate (longitudinal / lateral / combined tire forces) 3 Ability to predict ahead of vehicle (Yes / Possibly / No) 2 Detection of Gravel (Evaluation of system limits) 1 Availability of estimate (individual tire / per axle / global) 0 and highway maintenance facilities in large areas of Austria and Bavaria, especially in winter, [Ran13]. 2.2.4. Sensorfusion The idea of this application is to use as many different sensor signals as possible to calculate the most probable friction potential. One advantage is that the plausibility of individual signals can be checked by using the combination of other sensors, e.g. [KP09, p.77-80]. IntheFRICTI@Nprogramme, forexample, information fromdifferentexisting sensors thatmeasure thevehicle reaction, tire reactionandenvironmental characteristics were combined using sensorfusion and learning techniques, [KP09, p.10-11]. Ambient temperatures, infrared laser spectroscopy, laser scanner and different cameras were used for this purpose. These methods are expected to become even more important in the future. The involvement of vehicle-dynamics-based approaches in a sensorfusion system will provide a very valuable contribution to a robust and reliable estimate of the friction potential. Nevertheless, this work does not investigate sensorfusion approaches. 2.3. Requirements for a friction potential estimation method Lex et al. investigated the requirements for an estimation method using vehicle dynam- ics states to be applied in FCW, AEB and airbag pre-firing, [LEH11]. Nine criteria were presented, which were weighted based on their applicability to the aforementioned systems, see Table 2.3. Lex et al. used this criteria to preselect suitable methods out 36