Optimization of metro wheel reprofiling based on cluster analysis

Abstract

According to the characteristics of the metro vehicles running on the same track leading to a similar wheel wear, a clustering method based on k-means is proposed. The weighted wheel flange height of Sh and qR, and wheel flange thickness of Sd are used as clustering parameters. This paper classifies the treads of the wheels to be repaired and obtains a typical worn wheel profile to reprofile the wheels. Under the premise of ensuring the operation safety of vehicles, the wheel profile optimization model is established for a typical worn wheel profile based on the minimum repair volume. The optimized profile is used as the reprofiling template for all wheels in the same classification. Finally, based on the reprofiling standard for the wheel diameter difference between the left and right wheels, the optimization reprofiling of the wheel profile is achieved. The research results show that the proposed optimization strategy base on cluster analysis can significantly reduce the material removal.