基于故障预测的预防性维修策略优化模型

摘要

Aiming at a single-unit system subject to condition monitoring and its degradation state distribu-tion function is unknown, the fault prediction technology is introduced into the practice of preventive mainte-nance, puts forward a preventive maintenance strategy based on failure prediction. First, according to the re-maining life prediction results which are obtained by detecting state information, the residual life of preven-tive maintenance as threshold to draft preventive maintenance strategy. Then, according to the theory of re-newal process, establish an optimization model with preventive maintenance threshold and the prediction in-terval as optimization variables, the minimum average maintenance cost and maximum availability of equip-ment as optimization target function. By using Seeker Optimization Algorithm ( SOA) , the optimal preven-tive maintenance threshold and maintenance prediction intervals were obtained. Finally, using the known predicted results, the example is given to simulate optimization and get the best forecast cycle, in the guar-antee availability at the same time, get the minimum average maintenance cost of the equipment, verify the feasibility and effectiveness of the model.%针对具有状态检测且退化分布未知的单部件系统，把故障预测技术引入到预防性维修的实践中，提出一种基于故障预测的预防性维修策略。首先，根据由状态检测信息得到的剩余寿命的预测结果，以预防性维修时的剩余寿命为阀值制定预防性维修策略。然后，根据更新过程理论，建立以设备的预防性维修阀值和预测间隔期为优化变量，以设备平均维修费用最小和设备可用度最大为优化目标的预防性维修优化模型。采用人群搜索算法进行优化求解，得到系统最佳的预防性维修阀值和维修预测间隔期。最后，通过引入算例，对所建模型优化仿真求解，得到设备最佳的预测周期，在保证设备可用度的同时，使设备的平均维修费用最小，验证了所建模型的可行性和有效性。