The characterization of driver subjective evaluations is a major challenge in automotive industry, specifically in tire companies. A successful subjective/objective relationship requires vehicle response and the driver's evaluations to be simultaneously collected during closed-loop maneuvers. For this "unobtrusive observation" approach, the recognition of such driving situations is essential to identify key tire properties crucial for handling performance. The objective of this paper is two-fold. First, a novel algorithm for driving situations recognition is introduced. Based on signature analysis, this algorithm adapts to the patterns it recognizes, requiring more parameters for more complex driving situations. The performance of the algorithm is assessed in terms of misses and false positives rates with the help of test drivers. Second, the recognition accuracy is optimized using different cost functions and the results are validated, providing an optimal set of key parameters essential for successful driving situations recognition.
展开▼
