基于威布尔分布的可靠性评估——核电厂DCS模块的备件优化
Weibull-Based Reliability Assessment—Optimizing Spare Parts Inventory for NPP DCS Modules

分散控制系统是核电厂的神经中枢系统，其设备的可靠性对核电厂安全有直接影响。实际电厂DCS控制机柜卡件的运维经验表明，卡件在投运早期与损耗期的失效率将随时间变化，传统指数型寿命分布难以准确描述。本文采用三参数威布尔模型描述卡件寿命分布，采用相关系数优化法估计位置参数，采用最小二乘法估计尺度参数和形状参数。通过对比与经验可靠度的拟合情况，以及Kolmogorov-Smirnov (K-S)检验，表明了采用三参数威布尔寿命分布和参数估计方法的合理性。最后，基于保障度计算了某核电厂在一年内所需的FUM卡件的备件储备数量，并证明其优于备件储备定额导则的估计方法。研究结果可为核电厂的卡件可靠性评估与备件优化提供指导。
The distributed control system (DCS) is critical for nuclear power plants (NPPs) safety, with its modules’ reliability directly impacting plant operation. Failure data from operational NPPs reveal time-varying failure rates, with decreasing rates during early life and increasing rates in wear-out periods, challenging the conventional exponential distribution model. This study employs a three-parameter Weibull distribution to better characterize DCS module life distributions. The location parameter is estimated using correlation coefficient optimization. Scale and shape parameters are initially derived through least squares estimation. The proposed model’s efficacy is validated through two approaches: comparing reliability fitting curves with empirical reliability obtained by the Herd-Johnson method, and conducting Kolmogorov-Smirnov (K-S) tests for goodness-of-fit. The model is then applied to optimize spare parts inventory for FUM modules in an NPP, demonstrating improved accuracy over traditional guidelines. These findings provide valuable insights for enhancing DCS reliability assessment and spare parts management in NPPs.