单部件加速退化系统的视情维修策略优化

针对单部件加速退化系统,研究了基于灰色模型、偏最小二乘回归和改进灰狼算法的最优视情维修策略,通过优化检测间隔和临近失效阈值最小化系统的平均维修费用率。维修策略优化中,考虑到维修数据的稀疏特性,利用适合小样本建模的灰色模型理论建立系统的加速退化模型;考虑了维修次数、系统退化状态与系统维修用时之间的多重相关性,利用偏最小二乘法建立了维修用时的多变量回归模型。在此基础上,给出基于系统平均维修费用率的目标函数,利用改进灰狼算法求解最优决策变量。通过算例说明了该最优视情维修策略的可行性。

Optimization of condition-based maintenance strategy for single-unit accelerated degrading systems

Optimization of condition-based maintenance strategy is studied for single-unit accelerated degrading systems, to minimize the average maintenance cost rate, by setting the optimal inspection interval and the near-to-failure threshold. To handle the sparse maintenance data, the grey model theory is used to establish the accelerated degrading model of the system, as it is suitable for small-sample modeling. Considering that there exist multiple correlations among the number of previous repairs, the degradation state and the maintenance duration, the partial least square is used to establish a multivariable regression model. Then, the objective function based on average maintenance cost rate is given, and the improved grey wolf optimization method solves the optimal decision variables. Finally, a numerical example is given to illustrate the feasibility of the proposed optimization method of condition-based maintenance strategy.