Method development for analysis of customer driving data by using driving patterns and cluster algorithm

Methode zur Analyse von Kundenkollektivmessungen mittels Einzelfahrtsegmentierung und Clusteralgorithmen

Knowledge about representative use by customers is of great importance for customer orientated dimensioning and testing of innovative vehicle concepts with alternative drives. This information is crucial in order to fulfill customer specific requirements regarding vehicles with future mobility concepts like MaaS or TaaS. For this purpose 5. Mio of mileage was measured with Data-loggers by different customers of light commercial vehicles. Based on both customer data and literature references a method was developed to segment the data in trips and intermediate rests. Using a combination of principal component analysis, t-SNE method and K-Means cluster algorithm, the trips can be divided in separated groups. A similarity between trips from different customer groups can be seen in the result of the cluster analysis. Based on the clustered and parameterized trips and pauses, representative driving cycles can be created for different customer applications in a following step. These for example can be used in simulations for customer design of alternative drives.

Language

  • English
  • German

Media Info

  • Media Type: Web
  • Edition: 1st Edition
  • Features: References;
  • Pagination: pp 311-325
  • Monograph Title: Commercial Vehicle Technology 2020/2021: Proceedings of the 6th Commercial Vehicle Technology Symposium
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 01874255
  • Record Type: Publication
  • ISBN: 9783658297169
  • Files: TRIS
  • Created Date: Feb 23 2023 9:31AM