Optimization of Distribution Model for CNC Lathes Early Failures
To eliminate early failures of CNC lathes, it is necessary to study the early failure distribution to establish accurate early failure distribution model of CNC lathes. During the modeling process, the same group of failure data obeys Weibull distribution, exponential distribution, super-exponential distribution and gamma distribution at the same time. To optimize the failure distribution model, this paper used “relevant index” method by comparing the distribution function curve fitting and determined that this group of failure data fitted Weibull distribution using maximum likelihood estimation method with its shape parameter k 0.8706 and scale parameter b142.6991.
Fan Rui, Qiao Lihong, Chen Huawei, Ochi Akio, Usuki Hiroshi and Sekiya Katsuhiko
D. S. Chen and T. S. Lv, "Optimization of Distribution Model for CNC Lathes Early Failures", Key Engineering Materials, Vols. 407-408, pp. 107-111, 2009