Estimation of Parameters For A Dependent Bivariate Weibull Distribution

Consider a dependent bivariate Weibull model whose survival function is \barF\left(x_1, x_2\right)=\exp \left\-\left\left(x_1^1 / \alpha / \theta_1\right)^1 / \delta+\left(x_2^1 / \alpha / \theta_2\right)^1 / \delta\right^\delta\right\,with x_iθ_i＞0,（i=1,2） α＞0 and 0＜δ＜1 Based on the data of both components and series systems experiments under type I censoring, estimators of the unknown parameters θ₁,θ₂,αand δ are given, their asymptotic properties are discussed also. A simulation result is given, too.