基于性能数据的产品退化风险评估

风险优先数(RPN,Risk Priority Number)的风险因子由专家打分确定,主观性强。同时,评估过程依赖于失效数据,而高可靠性产品通常难以获取充足的失效数据进行风险评估。为了客观、有效地评估高可靠性产品的失效风险,提出了一种的改进RPN方法。根据退化数据建立产品的Gamma退化模型,推测退化失效的发生概率,评估退化失效模式的发生频度O;用非线性相关信息熵分析产品的性能数据,评估退化失效模式发生对产品功能的影响程度,以此衡量退化失效模式的严重度S;以发生频度和严重度的乘积作为RPN值衡量产品的退化风险。最后,以谐波齿轮为例,验证了该方法的正确性和有效性。

Product Degradation Risk Analysis Based on Performance Data

Theriskfactorof RPN method is evaluated by expert scoring which leads to subjective bias. And the evaluation is based on failuredata,but high reliability product is difficult to obtain sufficient failure data for risk evaluation. In order to evaluate the risk of high reliability product objectively and effectively,an improved RPN method was proposed. Gammadegradation model was established based on degradation data to estimate the failure rate,then the occurrence(O)of degradation failure was evaluated according to the failure rate. Severity(S)was defined as the influence of the degradation modes on the product function which was evaluated by analyzing the performance data using nonlinear correlation information entropy method. The RPN value was calculated by multiplying the occurrence and severity to measure the degradation risk. Finally,a case study of harmonic gear was performed,the result showed the effectiveness and accuracy of improved RPN method.