Abstract
Currently, the Remaining Useful Life (RUL) prediction accuracy of stochastic deterioration equipment is low. Existing researches did not consider the impact of imperfect maintenance on equipment degradation and maintenance decisions. Therefore, this paper proposed a remaining useful life prediction-based maintenance decision model under data-driven to extend equipment life, promoting sustainable development. The stochastic degradation model was established based on the nonlinear Wiener process. A combination of real-time update and offline estimation estimated the degradation model’s parameters and deduced the equipment’s RUL distribution. Based on the RUL prediction results, we established a maintenance decision model with the lowest long-term cost rate as the goal. Case analysis shows that the model proposed in this paper can improve the accuracy of RUL prediction and realize equipment sustainability.
Highlights
With the advancement of science and technology, the requirements for equipment reliability are getting higher and higher
With the rapid development of Sensor technology [5], Data mining [6], and Intelligent algorithms [7], Prognostics and Health Management (PHM) has gradually become the main management methods for complex equipment systems [8], which has attracted the attention of many scholars at home and abroad [9,10,11,12]
It is worth noting that M3 is lower than M2 in both MAE and Root Mean Square Error (RMSE), which indicates that the introduction of a non-linear Wiener process will make the prediction model more effective
Summary
With the advancement of science and technology, the requirements for equipment reliability are getting higher and higher. B. et al [43] developed an imperfect maintenance strategy for task-oriented systems, established a random equipment degradation model based on the Wiener degradation process. Wiener process and the homogeneous Poisson process to model the cumulative effect of equipment, thereby improving the accuracy of RUL prediction This model does not study equipment maintenance decisions and assumes that the number of imperfect repairs is unlimited. Chen, Y. et al [46] used nonlinear Wiener process and nonhomogeneous Poisson process based on literature [45] to establish an imperfect decisionmaking model that satisfies the upper limit of the number of repairs This model adopts the maximum likelihood method for parameter estimation and cannot update the parameters according to the equipment’s real-time data, making the predicted result inaccurate and affects the accuracy of maintenance decision. Taking the bearing as an example, we compared the validity and accuracy of the model
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have