非线性与线性平差偏差的分布特征

测量中的平差模型大都是非线性型的，实际处理常使其线性化，进行线性模型平差，平差值与其理论值必然存在差异。文中对非线性函数作任何逼近时所产生的参数偏差的概率分布特征进行讨论，导出了函数偏差与平差参数偏差的期望和方差公式，它们是进一步研究非线性平差统计分析的基础。

Deviation Distributive Features of Linear and Non-linear Adjustment

Adjustment models adopted in surveying are mostly non-linear. Linearization of models is needed in practice in order to carry out linear adjustment. So differences must exist between the adjusted value and the theoretical value. Probability distributive features of parameter deviation caused while approaching non-linear functions are discussed in the paper. Expectation and square deviation formulas for calculating function deviation and adjusted parameter deviation are derived. A foundation is laid for further study of non-linear adjustment statistical analyses.