Reliability allocation of industrial robot (IR) system is one of the important means to improve its whole life cycle, reduce maintenance cost, and characterize weak subsystems. The IR system is not only very complex but also has strong customization; meanwhile, its sample data are small, resulting in unclear degeneration and failure. Based on the above two epistemic uncertainties, a new methodology called multiple‐state IR system reliability allocation method with epistemic uncertainty (MIRS‐RAM‐EU) is proposed. First, the Dempster‐Shafer (D‐S) evidence theory is used to quantify the epistemic uncertainty. Then, the Kolmogorov differential equations of MIR's subsystems are calculated. The reliability index of MIRS is allocated based on Birnbaum importance degree theory, and the reliability allocation coefficient of each IR subsystem is clearly expressed by this method. Finally, compared with traditional importance allocation method, the MIRS‐RAM‐EU is more efficient and accurate. This method is usefully directive for reliability evaluation of IR.

中文翻译：

---

基于认知不确定性可靠性分配方法的多状态工业机器人系统研究

工业机器人（IR）系统的可靠性分配是提高其整个生命周期，降低维护成本以及表征弱子系统的重要手段之一。IR系统不仅非常复杂，而且具有强大的定制能力。同时，其样本数据很小，导致不清楚的退化和故障。基于以上两个认知不确定性，提出了一种具有不确定性的多状态红外系统可靠性分配方法（MIRS‐RAM‐EU）。首先，Dempster-Shafer（DS）证据理论用于量化认知不确定性。然后，计算了MIR子系统的Kolmogorov微分方程。MIRS的可靠性指标是根据Birnbaum重要度理论进行分配的，用这种方法可以清楚地表达每个红外子系统的可靠性分配系数。最后，与传统的重要性分配方法相比，MIRS-RAM-EU更加有效和准确。该方法对于IR的可靠性评估很有用。