现有结构的荷载概率模型及其统计推断

Probability model and its statistical inference for loads on existing structure

现有结构上的荷载并不能简单沿用设计阶段的概率模型，它的概率模型应根据现实、特定结构和环境的历史与当前信息重新建立，需要解决小样本条件下的统计推断问题．本文讨论了现有结构上荷载的随机性和概率模型转换问题，通过综合利用自然信息和意识信息，提出在小样本条件下推断荷载概率分布函数和模型参数的方法，为建立现有结构上荷载的概率模型提供了系统的方法，可供制定有关标准参考．

Instead of adopting probability models set up in design stage, probablility models for loads on existing structure should be reset up by using the information of realistic and specific structure and its environment, which calls for the solution to the problem statistically inferred with small samples. Randomness of loads on existing structure and transformation of their probability models are discussed in this paper, and methods to statistically infer parameters of probability models by combining natural information and consciousness information are proposed. These give a series of methods for setting up probability models of loads on existing structure with small samples, as well as reference for calibration.