模糊综合评价的非离散抽样模型

Non discrete sample model for fuzzy comprehensive evaluation

为对工业产品质量状况进行综合评价 ,提出了一种以模糊数学为基础的非离散抽样理论。解决这类问题的办法是用样本所有个体的合成指标值来推断总体中相应指标值的等级隶属度。定量计算公式可以通过离散与非离散模型之间的变量映射 ,由离散模型导出。另外 ,借助于间接估计指标分子和分母置信区间的办法 ,克服了估计指标置信区间的困难。

A non discrete sample theory was developed using fuzzy mathematics for the comprehensive evaluation of industrial product quality. The compositive index value of all individual in the sample is used to deduce the classmembership of the corresponding index value in the general aggregate. Quantitative formulas result from the discrete sample model through the mapping of variables between both models. In addition, the difficulty of estimating the index credible interval was overcome by an indirect method which estimated the credible intervals for the numerator and denominator of the index. The model has applied to study the quality of a thousand key enterprises throughout the nation to facilitate coming management reform.