发射场非标设备失效概率获取方法研究

Research on Measuring the Failure Probability of Launch Sites' Non-Standard Equipment

针对发射场概率风险评估中非标设备失效概率难以获取的问题,提出了建立非标设备失效故障树;用Delphi法确定基本事件中部件初始失效概率;用短期运行数据结合Bayes方法对部件初始失效概率加以修正,由故障树解算获取非标设备失效概率,即Delphi-Bayes法。以某厂房桥式起重机总电源不能切断故障为例,应用Delphi-Bayes法确定了故障树基本事件中部件失效概率,获取了起重机总电源不能切断的失效概率;以疲劳加速试验数据为标准,计算了所获取部件及设备失效概率的偏差,其偏差在可接受范围内。经验证,Delphi-Bayes法切实可行,所得非标设备失效概率能够满足发射场定量概率风险评估的要求。

In order to solve the difficulties in the acquisition of the failure probability of non-standard equipment during a launch site PSA, Delphi-Bayes Method was put forward in this article. Firstly, the Fault Tree of non-standard equipment failure was established. Secondly, the initial probabilities of components＇ failures by Delphi Method were calculated. Then the initial probabilities were amended combining the short-term service data with the Bayes Method. At last, the failure probability of non-standard equipment was worked out using the Fault Tree. Taking a bridge crane in some factory for example, the components＇ failure probabilities in the basic events of the Fault Tree were determined and the failure probability of a crane whose general power cannot be cut off was obtained by applying the Delphi-Bayes Method. Additionally, according to the Fatigue Acceleration＇ s experimental statistics, the deviation of the failure probabili- ties of acquired components and equipment was also worked out, the result was acceptable. Delphi-Bayes Method proves to be practical and the acquired failure probability of the non-standard equipment satisfies the requirements of the launch sites＇ quantitative PSA.