Driving Cycle Analysis Methods Using Data Clustering for Machine Design Optimization

Abstract

Machine design optimization is usually based on one or few operating points, but when a driving cycle is considered, cluster analysis can be performed to determine the representative operating points of the driving cycle, which can be used in the optimization process to enhance the overall efficiency of the driving cycle. Furthermore, the machine performance indices can be guaranteed in different domains, especially if multiphysics design optimization is considered. This work discusses the basics of analyzing the driving cycle to determine the clusters, representative points and optimization weights. The paper weighs up the pros and cons of the data clustering methods for their application in machine design optimization. This paper proposes the X-Means method to determine the clusters and representative points of the torque-speed profile. Moreover, a case study is performed on the WLTP driving cycle to evaluate the effectiveness of the proposed method for driving cycle analysis for machine design optimization.