针对三参数威布尔分布在参数估计中存在的问题，首先提出改进的整体最小二乘法(ITLS)进行参数估计，然后采用差分进化算法(DE)进行参数优化． DE 在优化过程中可以准确、快速地寻得目标函数全局最优解;改进整体最小二乘法在回归分析时兼顾自变量和因变量误差，使分析结果更加真实合理，以偏差平方和作为约束准则进行优化，使所得曲线拟合效果达到最佳并且更加符合实际情况． 最后应用文中所讨论的方法建立了某型号加工中心三参数威布尔分布的可靠性模型，并对结果进行对比分析． 分析表明所提出的方法可以有效提高加工中心可靠性建模的准确性和真实性，从而为整机可靠性建模提供有效的建模方法．

In order to solve the problems existing in the parameter estimation of three-parameter Weibull distribu-tion，an improved total least square (ITLS) method is proposed for parameter estimation，and a differential evolu-tion (DE) algorithm is adopted for parameter optimization． DE helps to accurately and rapidly find a global optimal solution of the objective function in the optimization process，and ITLS helps to obtain realistic and reasonable re-sults because it considers not only the independent but also the dependent variables during the regression analysis．At the same time，by taking the square sum of deviations as the constraint rule，ITLS helps to achieve the best curve fitting effect with higher authenticity． The proposed method is then used to establish a three-parameter Weibull distribution model of a machining center，and the results are compared and analyzed． Analytical results show that the proposed method effectively improves the accuracy and authenticity of the reliability model of machi-ning centers and thus provides an effective method for the reliability modeling of the complete appliance．