幂律过程拟合优度检验的功效比较

Power Comparisons of Goodness of Fit Tests for Power Law Process

幂律过程常被用于描述复杂系统的可靠性增长模型和可修系统的可靠性分析。本文考虑定数截尾场合幂律过程的７种拟合优度检验统计量，在备选假设为对数幂律过程时，比较了这７种检验法的功效。大量随机模拟计算表明，当ｒ在较宽的范围内，Ｆ检验法的功效比其余６种检验法都要高。

The power law process is often used as a model for reliability growth of complex systems or reliablity of repairable systems. In this paper, seven statistics are consi dered for a goodness of fit tests of a power law process in the case of failure truncated data. The results of a power comparison are given under the alternative hypothesis that failure data came from a log power law process. Monte Carlo simulations shows that the F statisitcs has the highest power when r is in the some wider region.