This paper proposes an evolutionary fuzzy system for fault diagnosis of battery packs used in electric vehicles. Based on the knowledge of battery and experience of engineers in the field of electric vehicles, input and output variables are carefully selected in the proposed system to represent the characteristics of the battery packs. In the design process, a set of basic (primitive) expert rules is defined by battery engineers and is used for the initialization of the evolutionary process. A typical three-component fuzzy system paradigm is chosen that consists of fuzzification, inference engine with fuzzy rules, and defuzzification. In this paper, a genetic algorithm is used to optimize the proposed fuzzy system. This design process can simultaneously evolve the fuzzy membership functions, fuzzy rule set and weights of each rule based on input-output training data. A case study of the diagnostic system design for lead-acid battery packs is presented. The simulation results show the optimized fuzzy diagnostic system produces very good outputs that match the target data collected from the vehicle test. The evolved fuzzy system is much more effective in comparison to the fuzzy system defined and tuned by the battery experts. The method can be applied to other battery chemistries and energy storage devices.
Fuzzy diagnostic system for EV traction battery pack
Fuzzy gestüztes Diagnosesystem für das Batteriesystem eines Elektrofahrzeugs
2001
11 Seiten, 5 Bilder, 5 Quellen
Aufsatz (Konferenz)
Datenträger
Englisch
Fuzzy Diagnostic System for EV Traction Battery Pack
British Library Conference Proceedings | 2001
|Fuzzy diagnostic system for EV traction battery pack
Kraftfahrwesen | 2001
|Fuzzy Diagnostic System for EV Traction Battery Pack
British Library Conference Proceedings | 2001
|