Real-time remaining useful life prediction based on adaptive kernel window width density

Abstract

Remaining useful life (RUL) prediction plays an important role in improving the availability and productivity of systems. To improve the accuracy of real-time remaining useful life prediction during system operation, we propose a modeling method for real-time remaining useful life prediction based on adaptive kernel window width density. Firstly, a non-parametric kernel density estimation real-time remaining useful life prediction model is proposed and a window width model with adaptive kernel window width density is established by introducing a local density factor in the window width selection. The local density of sample points is calculated by the k-nearest neighbor distance, and the kernel density estimation is performed by adaptively selecting the window width value according to the local density of sample points in the region of nonuniform distribution of sample points. As the monitoring data changes in real time, the kernel density estimates of known samples are used to recursively update the kernel density estimates of new samples. Moreover, the logarithmic transformation of random variables and space mapping are used in the establishment of the remaining useful life prediction model. The model of logarithmic kernel diffeomorphism transformation is established to solve the boundary shift problem of kernel estimation in the prediction for improving the prediction accuracy. Finally, the validity of the method is verified through case studies and the accuracy of the model is judged using evaluation quasi-measures.

Multiscale global and local self-attention-based network for remaining useful life prediction

“…Figure 11(d) shows the attention weights learned from the local feature extracted network, where the weight value at row i and column j represents the contributions of the ith local field to the j field. We 1 , where the score will be generated. We applied the same model (MGLSN) that we used in the C-MAPSS experiments with a time window of 60 to the PHM-2008 dataset.…”

“…Prognostics and health management (PHM) is an important technology that ensures the operational reliability of large * Authors to whom any correspondence should be addressed. modern equipment and reduces redundant engineering maintenance and costs in many industrial areas [1,2]. The goals of PHM are to improve the system reliability and safety by monitoring various industrial sensor data measured from the equipment [3,4].…”

Performance and Challenges of Planetary Roller Screw Mechanism in Complex Conditions: A Comprehensive Review of Recent Progress

“…The advancements in ball screw diagnostics are exemplified by Zhang et al [146] through an instance-based ensemble deep transfer learning network (IEDT), which leverages auxiliary data from related domains to overcome the challenge of limited labeled data in target domains. The integration of preload and precision parameters into a novel index for wear degradation prediction by Zhang et al [147] and the systematic methodology for ball screw prognosability by Li et al [148] underscore the criticality of early diagnostics in machinery maintenance. Techniques developed by Han et al [149] for long-range running ball screw machines emphasize the significance of monitoring lubrication conditions to forestall premature failures.…”

Real-time remaining useful life kernel density estimation considering dynamic transition of degradation states

“…In this section, to verify that the gear teeth are in contact with each other under the action of surface pressure and sliding friction during the periodic meshing process, the test adopts a power flow closed test bench as shown in figure 2, The test bed is composed of a main test gear box and an auxiliary gear box with a center distance of 150 mm [13]. The test is loaded with a mechanical lever, and the torque is measured with a torque-speed sensor.…”

“…Wang H et al [12] considering the joint dependency of degradation rate and variation on time-varying operating conditions. Zhang J et al [13] proposed a modeling method for real-time RUL prediction based on adaptive kernel window width density. Diyin T et al [14] proposed a Markov process-based RUL prediction method for estimating the real-time average RUL of degraded systems whose states are continuously monitored under dynamic operating conditions.…”