1.North Central University;2.Taiyuan University of Science and Technology
The National Natural Science Foundation of China (General Program, Key Program, Major Research Plan)
随着检测传感技术的发展,诸如风力发电机叶片等可对其状态进行检测,并依据检测结果进行剩余寿命预 测.但此类系统在运行中受环境冲击影响较大,如雷击对叶片裂纹的影响.如何对冲击影响下的系统剩余寿命进行 预测,并结合预测结果进行经济可靠的维修决策是一个值得研究的问题.本文针对状态可检测的连续退化系统,研 究了考虑加速冲击损伤特性下的系统剩余寿命预测及基于预测的维修决策.首先,考虑了自然退化和与退化相关 的冲击损伤,构建了加速冲击损伤退化模型和剩余寿命预测模型;其次,制定了基于周期检测的状态维修与预测维 修相结合的混合维修策略,并推导了不同维修活动的发生概率;然后,构建了以长期平均费用率最小为目标,以检测 间隔和故障率阈值为决策变量的决策模型,并给出了优化解法.最后,以风力发电机叶片为案例验证了模型的适用 性和有效性,对系统的参数进行了灵敏度分析,并与未考虑加速冲击损伤和未考虑预测的维修决策结果进行了对 比分析.
With the development of detection and sensing technology, status of systems such as wind turbine blades can be detected, and Remaining Useful Life (RUL) can be predicted according to the detection results. However, such systems are greatly impacted by environmental shock in operation, such as the shock of lightning on blade cracks. It is worth studying how to predict the RUL of systems under impact and how to make economic and reliable maintenance decisions link the prediction results. In this paper, the RUL prediction and the Predictive Maintenance (PdM) decision are studied for the continuous deteriorating system with condition detectability. Firstly, the accelerated shock damage degradation model and the RUL prediction model are modeled considering the degradation and the degradation-related shock damage. Secondly, the combined maintenance policy of Condition Based Maintenance (CBM) and PdM on periodic inspection is formulated, and the occurrence probability of different maintenance activities is derived. Then, the maintenance decision model is constructed with the objective of minimizing the long-term average cost rate and the decision variables of inspection period and failure rate threshold, and the optimized solution is given. Finally, the applicability and effectiveness of the model are verified by taking the fan turbine blade as an example, and the sensitivity analysis of the system parameters is carried out. The optimized solution is compared with the maintenance decision results without accelerated shock damage and without prediction.