发电机组预知性维修管理算法

Study on Algorithm of Management of Power Plant Maintenance

提出了以可靠性理论及机组、设备历史情况为基础，通过计算设备最大可能故障时间和故障时间百分比置信区间，预测未来设备故障情况，指导维修人员进行预知性维修管理的算法。为了能考虑各种计划或非计划检修对设备状态的影响，引入了可以用来描述检修效果的年龄减轻因子，对参数计算结果进行修正。通过预测结果和实际运行记录的对比，证明该算法是可行的。

Based on the reliability theory and equipment's historical records, this paper presents a method to forecast the failure of generator units in the future by calculating the failure time of maximum probability and the left side value of failure time of percentage section. It may help staffs to conduct the predictable maintenace. In order to eliminate the effect of repairs in the past, an age reduction factor is introduced to revise the result. This method is proved feasible by comparing calculated result of a specific plant with its operation diary.