An Enhanced Copula-Based Method for Battery Capacity Prognosis Considering Insufficient Training Data Sets

摘要

Data-driven based prognostics typically requires sufficient run-to-failure training data in order to gain knowledge of the degradation characteristic of engineering components or products without understanding the fundamental degradation mechanisms. With insufficient training data, however, the model learned from the training data may be inaccurate, which could result in large prediction errors for the remaining useful life (RUL) of many actual test units. This paper proposes an enhanced copula-based prognosis method to address the limitations for insufficient training data. Effectiveness of the proposed method is demonstrated using the capacity degradation data from four lithium-ion batteries.