A key question in the development of a tire is "How can this tire improve vehicle handling?" Good handling tires contribute not only to active safety of vehicles, but also to the pleasure of driving. Handling performance is largely determined by the driver. Therefore, the final and most important handling assessment of tires is done by professional test drivers driving on a handling circuit and giving their subjective opinion. This provides the tire manufacturer with the important tire handling performance, but it gives limited information on what the driver perceives as good and how his opinion is formed; the driver is still a 'black box'. Three methods, all based on field experiments, for gaining this knowledge about subjective assessments were chosen for this research. They have in common that they predict the driver's subjective assessment of tire handling, based on vehicle dynamics measurements. The differences lie in the way they derive and utilize these measurements. For method 1, the prediction is done with a General Regression Neural Network based on vehicle dynamics measurements. With this method, several limitations for using regression for tire handling can be circumvented.Method 2 focuses on the driver's workload as an indication for his subjective assessment. This method derives from the fact that the driver adapts to changing vehicle handling behavior. Method 3 also focuses on the driver but not by looking at measures from 'outside' the driver, like workload measures, but by modeling the driver behavior during closed-loop driver-vehicle simulations and looking at driver parameters 'inside' the driver (model). The results show that all three methods can predict the driver's opinion about tire handling, based on vehicle dynamics measures. Analysis of the relevant measures for the prediction of methods 1 and 2, provides information on the 'what'-question. Likewise, method 3 provides information on the 'how'-question. In addition, drivers adapting behavior, e.g., compensating for less good handling tires by investing more effort, can be quantified with the mental workload measures. This makes them good indicators of driver's perceived tire handling behavior, even when the performance measures do not show differences. For implementation of one or more methods, only a subset of the vehicle dynamics measurements used for this research is needed. During use, the methods can be adapted to changing tire testing methods, with different measurements, handling aspects or maneuvers. When vehicle and tire models are available, these methods can also be used for virtual testing, predicting driver's opinion on tire handling. This research provides a first step in opening up the 'black box' of the driver by quantifying the driver's tire feeling.
|Translated title of the contribution||Feel the Tire - Tire Influence on Driver’s Handling Assessment|
|Publication status||Published - 2015|
|MoE publication type||G4 Doctoral dissertation (monograph)|
- tire characteristics
- driver modelling
- driver mental workload
- general regression neural network