A Linearization Regression Method to Determine Life Data Distribution Type

Abstract

The law of equipments or products life is the calculation fundament of reliability indexes in power plant. Aim to how to determine the most fitted life distribution type, the linear regression and correlation coefficient had been employed in this paper. The linearization of common distribution like as: exponential distribution, normal distribution and logarithmic normal distribution, is complied with the principle of linear regression. After the comparison of correlation coefficient, the most fitted distribution type is quickly found out. The result also shows that the inferred distribution type is in accord with the facts. We may draw the conclusion from this research that the optimized linear regression and correlation coefficient is effective and valid for the inference of life data distribution type.