中文摘要：

考虑包含如方差分量模型、似乎不相关回归方程模型、增长曲线模型和扩充的增长曲线模型等众多常见模型的一类较广泛的线性模型。对模型中误差向量的分布不作假定时,给出了在二次损失或矩阵损失下存在回归系数的线性可估函数的一致最小风险线性无偏估计的充分必要条件。

英文摘要：

This paper studies a general class of linear models including the variance com- ponents model, the seemingly unrelated regression equations model, the growth curve model, the extended growth curve model, and so on. Without assuming the concrete distribution of the model error, the necessary and sufficient existence conditions are established for uniformly min- imum risk linear unbiased estimators of the estimable linear functions of regression coefficients under quadratic losses and matrix losses~ respectively.