Dual-response optimization for lifetime improvement experimental designs

Framework for robust parameter design and optimization of reliability characteristics

“…Thus, warranty should be considered to improve reliability. Wang et al 33 . constructed the optimization criteria for reliability improvement by introducing warranty ( y 0 ), and proved the maximization of the reliability function to be equal to minimizing $$\frac{{{y_0} - {y_p}}}{\sigma }$$.…”

“…Thus, in most previous reliability improvement studies, only the main effects and a few two-factor interaction effects have been considered. 32,33 However, without fully considering significant interactions, the reliability may be overestimated, and the corresponding optimal factor settings may be inappropriate . 34 To resolve this problem, a bootstrap-based methodology was introduced to identify significant variables.…”

“…However, without fully considering significant interactions, the reliability may be overestimated, and the corresponding optimal factor settings may be inappropriate 34 . To resolve this problem, a bootstrap‐based methodology was introduced to identify significant variables 35,33 . In addition, Beattie et al 36 .…”

“…34 To resolve this problem, a bootstrap-based methodology was introduced to identify significant variables. 35,33 In addition, Beattie et al 36 developed a two-stage Bayesian variable selection method to identify significant variables. Wang and Ma 37 proposed a Bayesian variable selection method that combined generalized models with factorial effect principles.…”

“…Restricted by degree of freedom constraints, however, only a few effects can be included in the traditional model. Thus, in most previous reliability improvement studies, only the main effects and a few two‐factor interaction effects have been considered 32,33 . However, without fully considering significant interactions, the reliability may be overestimated, and the corresponding optimal factor settings may be inappropriate 34 .…”