Construction of new energy vehicle maintenance decision support system based on knowledge graph
Year: 2026
Type: conference-paper
Abstract: Aiming at the core problems in the process of new energy vehicle maintenance, such as low efficiency of fault diagnosis, difficulty of cross-system knowledge integration, high dependence on individual experience in the decision-making process, and insufficient utilization of historical data, this paper proposes a new energy vehicle maintenance decision support system that deeply integrates knowledge graph and rule-based reasoning technology. First, the system combing new energy vehicles SanDian system (battery, motor and electric control) and the vehicle structure characteristics and typical failure mode of the controller, to build a covered entity, failure mode, symptoms, detection methods, maintenance plan, case knowledge such as multidimensional associated maintenance knowledge map. Second, design a hierarchical hybrid reasoning mechanism, combining rule-based rapid diagnosis and based on the graph embedding semantic association mining, from complex, fuzzy symptom to accurate traceability and intelligent positioning of potential root causes. Further, the system integration service resources, cost, difficulty, such as multi-objective optimization model, to identify the fault generated hierarchical, personalized service strategy recommendations. Finally, the diagnostic accuracy, decision-making efficiency and knowledge coverage completeness of the system are quantitatively verified by constructing a measured case base containing multiple models and multiple fault scenarios. Results show that the built system can effectively integrate fragmentation maintenance knowledge, significantly increased complex fault diagnosis speed and accuracy, and reduce the absolute dependence of expert experience, to establish a standardized and intelligent new energy car after market service system to provide key technical support.
Author: Bo Wang
Institution: Liaoning Vocational University of Technology
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7205015484