Evaluation Model for Vehicle Shift Quality Based on Evidence Theory and FNN

Abstract

To achieve the shift quality evaluation effectively, reasonably and timely, a new model to evaluate shift quality based on evidence theory and fuzzy neural network (FNN) is proposed in the paper. A method combining subjective evaluation and objective evaluation is adopted. Subjective evaluation provided by various drivers is aggregated by means of evidence theory, which can improve the reliability of subjective evaluation. Objective evaluation metrics measured by instrument and the corresponding subjective evaluation are self-learned and trained with FNN. Shift quality evaluation system is established, which is used to convert evaluation metric to subjective rating. And the correlation between the developed evaluation model and subjective ratings given by drivers is proved. Computational and experimental results show that the proposed model is available and promising. The proposed model is a theory base for developing control strategies and improving shift quality.