论文部分内容阅读
针对现有故障预测算法性能评估指标受实际剩余使用寿命约束的问题,从稳定性角度提出一种评估算法性能的方法.通过研究对象系统健康退化过程,在对象系统实际剩余使用寿命未知情况下,利用可以实时获得的剩余使用寿命预测值和已消耗寿命值,通过计算虚构寿命值的变异系数指标来客观评估故障预测算法的性能.为了验证所提方法的有效性,结合机电作动器故障演化模型仿真生成数据对递归最小二乘和粒子滤波两种故障预测算法的稳定性进行了实时评价.仿真结果表明,所提方法与运用已有指标、在获知剩余使用寿命理想值前提下得出的评估结果保持一致.
Aiming at the problem that the performance evaluation index of the existing fault prediction algorithm is constrained by the actual remaining life, a method of evaluating the performance of the algorithm is proposed from the perspective of stability. By studying the health degradation process of the target system, when the actual remaining life of the target system is unknown, The objective function of the fault prediction algorithm is to evaluate objectively the performance of the fault prediction algorithm by calculating the residual life expectancy value and the life expectancy value obtained in real time.In order to verify the effectiveness of the proposed method, The stability of two kinds of fault prediction algorithms such as recursive least squares and particle filter are evaluated in real time by simulation data.The simulation results show that the proposed method and the application of the existing indicators have been obtained on the premise of knowing the ideal value of remaining life expectancy The assessment results are consistent.