Estimation of the lifetime distribution of mechatronic systems in the presence of a covariate: A comparison among parametric, semiparametric and nonparametric models

Bobrowski, Sebastian; Chen, Hong; Döring, Maik; Jensen, Uwe; Schinköthe, Wolfgang

doi:10.1016/j.ress.2015.02.012

Cited by 8 publications

(7 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…According to equations (6) and (18), the GEV proportional hazard function, with k = 1 and t f = 16 mn, has the final form…”

Section: Gev Phm For Cutting Toolmentioning

confidence: 99%

“…3,4 The flexible model which can be used to separate the effects of explanatory variables in such a context is the PHM introduced by Cox. 5 This model assumes that explanatory variables or operating conditions have multiplicative rather than additive effects on the hazard rate (see Bobrowski et al 6 ). In the Cox model, when the baseline hazard function is specified, we obtain the parametric PHM.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Proportional hazard model for cutting tool recovery in machining

Diamoutene

Noureddine

et al. 2019

Proceedings of the Institution of Mechanical Engineers, Part O:

View full text Add to dashboard Cite

The proportional hazard model is a statistical method capable of including information on environmental and operating conditions. In machining, in the reliability field of a cutting tool, the interest of using proportional hazard model is highlighted. On one hand, the environmental and operating conditions are described and taken into account as explanatory variables. Three covariates are considered, namely, the vibration signal, the hardness material, and the lubrication/cooling. On the other hand, a new baseline hazard function is designed according to phenomena of tiny tool breakage followed by a self-sharpening process. This latter phenomenon, which can be considered as a rare event, prompted us to study extreme value theory to propose an adequate baseline hazard function. The newly obtained baseline hazard function will be named generalized extreme value proportional hazard model. This function is obtained thanks to the Gumbel function and has the property to be non-monotonic, an increasing then decreasing function. An alternative option as a baseline hazard function, based on the flexible Weibull distributions, is also proposed. Results produced in this article show the impact of all these variables on the surface roughness of the machined parts. According to reliability studies, the premature replacement of the cutting tool implying financial losses can be delayed. This may be of particular significance and benefit, in terms of sustainable development, in the case of mass production, by limiting the frequency of tool replacement.

show abstract

“…According to equations (6) and (18), the GEV proportional hazard function, with k = 1 and t f = 16 mn, has the final form…”

Section: Gev Phm For Cutting Toolmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Proportional hazard model for cutting tool recovery in machining

Diamoutene

Noureddine

et al. 2019

Proceedings of the Institution of Mechanical Engineers, Part O:

View full text Add to dashboard Cite

show abstract

“…[14][15][16][17] Several procedures have been proposed for the reliability modeling of repairable systems. [18][19][20][21][22][23][24][25][26][27][28][29][30] Sehgal et al 20 suggested a procedure for the evaluation and selection of components such as rolling contact bearings for a given application considering reliability aspect, which is based on graph theory and matrix method. Pulido et al 21 presented a comprehensive methodology using Bayesian approach for obtaining prior distribution in reliability and require only prior intervals for the mean and the standard deviations of TTF of the product.…”

Section: Literature Reviewmentioning

confidence: 99%

Selection of time-to-failure model for computerized numerical control turning center based on the assessment of trends in maintenance data

Patil

Kothavale

Waghmode³

2018

Proceedings of the Institution of Mechanical Engineers, Part O:

View full text Add to dashboard Cite

This article provides a generalized framework for selection of time-to-failure model based on the assessment of trends in failure and repair time data. This framework is based on modifications of existing frameworks and can be applied for binary as well as multi-state systems. The proposed framework is applied for reliability analysis of a computerized numerical control turning center. For analysis purpose, the failure data are collected for 50 computerized numerical control turning center over a period of 7 years for three different working conditions, that is, when machining material is steel, aluminum and cast iron. The data collected are then processed using the proposed framework and the best-fit distribution is found for the time-to-failure data. Furthermore, the reliable life and reliabilities of the different sub-systems are estimated. From the analysis, it is found that spindle system, computerized numerical control system, electrical and electronic system, hydraulic system and cooling system are found to be critical from reliability and maintainability point of view. The analysis presented here is expected to help the users and manufacturers of computerized numerical control turning center to estimate the reliability in accurate manner.

show abstract

“…7 shows the results of estimation of the trend function by using the pilot estimator (6) and the estimator obtained by Algorithm 2. In the bottom panel we present the estimated renewal hazard.…”

Section: Examples With Real Datamentioning

confidence: 99%

“…For example, Gamiz and Roman [12]; Xiao and Dohi [30] and Taylor and Pena [29] use nonparametric methods in applications related to repairable systems, while Gandy and Jensen [15]; Luo et al [24]; Bobrowski et al [6]; Zhao et al [32] use kernel smoothing methods in specific reliability applications.…”

Section: Introductionmentioning

confidence: 99%