引言 观测数据的性质决定了所用的处理方法。自然界和科学实验中,有许多观测得不到确定的结果,这种数据在性质上就是随机的。地震前兆测量结果目前就属于这种情况,在分析处理这种数据时,充分运用现代研究随机现象的概率统计工具是必要的。 实测资料和测量系统性能的分析表明,前兆观测结果是一个多维随机向量的时间序 Introduction The nature of the observed data determines the approach used. In nature and in scientific experiments, many observations do not give definitive results, and the data are random in nature. Seismic precursory measurements now fall into this category, and it is necessary to make full use of the statistical tools of modern research on stochastic phenomena in the analysis and processing of such data. The analysis of measured data and measurement system performance shows that the result of precursor observation is the time series of a multidimensional random vector