摘要
图解法是求解Ｗｅｉｂｕｌ分布中３个可调参数的传统方法。但参数的推求过程要图解试求，反复调整，最后结果往往不是最佳拟合参数；线性最小二乘法是最近几年提出的一种求解Ｗｅｉｂｕｌ分布参数的方法，但其精度受位置参数初始迭代值的影响很大。文中运用麦奎尔特法推求Ｗｅｉｂｕｌ分布的参数，实现了参数推求的一举寻优。算例表明，拟合方法可行实用，拟合精度明显提高。
The graphic method is a conventional method to determine the three adjustable parameters of Weibull distribution. But the trial and error method is usually needed and the results are not the best ones. The linear least square method is proposed to obtain the parameters in recent years. But the solution accuracy is influenced by the initial integrative values of the parameters. In this paper a direct way based on the Marquardt method is developed to obtain the best fitting parameters. The examples show that this new method is feasible and can greatly increase the fitting accuracy.
