本文用1989年秋冬两季总悬浮微粒(TSP)样品及交通、工业和燃煤三种污染源样品的12种组份的数据,采用以PCA为主,结合理论概率方法解决了主成分的指向分析和污染来源的结构分析。用统计学方法解决指标指向与监测指标筛选问题。对现有监测点作了肯定评价。同时对有关问题进行了讨论。 In this paper, we use the data of 12 kinds of components of total aerosol samples (TSP) and three types of pollutants from transportation, industry and coal in the fall and winter of 1989, and use PCA as main method to solve the principal component analysis And the structural analysis of sources of pollution. Solve the problem of index pointing and monitoring index screening with statistical methods. The existing monitoring points were affirmed. At the same time, the relevant issues were discussed.