楼面活荷载标准值和设计值的小样本推断方法

SMALL SAMPLE METHOD FOR INFERRING CHARACTERISTIC AND DESIGN VALUE OF FLOOR LIVE LOADS

工程实际中有时需在测试数据不足的条件下推断楼面活荷载的标准值和设计值,这时的推断结果受统计不定性的影响较大,目前方法的推断结果偏于冒进,需建立小样本的推断方法。首先根据极大值I型分布参数、分位值的最好线性无偏和不变估计,提出了楼面活荷载标准值和设计值的线性回归推断方法;其次,根据贝叶斯理论,提出了楼面活荷载标准值和设计值的贝叶斯推断方法,并且对比分析了上述两种小样本推断方法的优劣,解决了应用中的具体问题,为小样本条件下楼面活荷载标准值和设计值的推断提供了实用的方法。

It is necessary in some cases to infer the characteristic and design values of floor live loads in condition of Small samples, where the inference results are usually influenced remarkably by the statistical uncertainty.Since the present methods generally lead to unsafe results, the small sample inference method is expected.By using the best linear unbiased and invariant estimation of the parameters and fractile of largest extreme value distribution typeⅠ, it we firstly proposed a small sample method for inferring the characteristic and design values of floor live loads, called linear regression estimation.According to the Bayes theory, it was proposed the Bayesian method for inferring characteristic and design value of floor live loads.Then, we contrasted the above two kinds of small sample inference method and solved some specific problems in the application.Which provides a practical method for inferring characteristic and design value of floor live loads in condition of Small samples.