作为光纤陀螺仪中的关键元件与薄弱环节，超辐射发光二极管(SLD)的可靠性在很大程度上决定了光纤陀螺仪的可靠性。针对其长寿命特点，研究了基于性能退化数据的可靠性评估方法。在对SLD失效机理分析的基础上，提出用正态泊松复合随机过程模型对产品在环境应力作用下的退化特性进行建模，基于所得模型，由SLD的性能退化信息估计模型中的参数进而评估得到SLD可靠性指标。克服了传统可靠性分析方法依赖寿命数据的缺点，能够在没有寿命数据的情况下评估得到SLD的可靠性指标，从而节约大量的试验经费和时间。

As the most important and fragile part of fiber optic gyroscopes (FOG), the reliabiliey of superluminescent diodes (SLD) determines the operational reliability and maintenance expenses of FOGs. Reliability of SLDs is studied based on degradation data for their long life characteristics. Based on analysis of degradation mechanisms of SLDs under environmental stress firstly, normal-poisson hybrid stochastic process model is presented as their life distribution model whose parameters can be estimated from the degradation measures of the SLDs. The proposed method can obtain SLDs′ reliability without failure data, therefore overcomes the drawbacks of traditional time-to-failure analysis method which requires failure data. This estimation method shows its significance in test costs and time.