A comparative study of statistical and soft computing techniques for reliability prediction of automotive manufacturing

2020 
Abstract Reliability and safety analyses are the most important activities for reducing risk of failure events and upgrading availability of manufacturing industries. The traditional statistical models have been currently used; however, the complexity growth and diversity of systems as well as uncertainty of their functions result in extreme difficulties in analyzing the reliability by such models. To overcome such drawbacks, the soft computing techniques are useful alternative for modeling of complex systems and prediction applications. Hence, this paper provides a comparative structure for predicting the operational reliability in automotive manufacturing industry, using soft computing + statistical techniques. The results of comparative structure revealed that the soft computing techniques can estimate the reliability function with the lowest error in all cases. Based on the performance criteria, it was observed that among the soft computing techniques, the Adaptive Neuro-Fuzzy Inference System (ANFIS) model yields better results in most cases and thus can be used for predicting operational reliability, since it predicts the reliability more accurately and precisely than the statistical models. Ultimately, the maintenance intervals based on the ANFIS model are proposed to upgrade the reliability and safety of automotive manufacturing process.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    62
    References
    9
    Citations
    NaN
    KQI
    []