摘要
本文综合利用统计和模糊识别方法处理了塘18井天然气逸出量连续7年的观测资料。所用的统计方法是方差分析,模糊方法则为模糊模式识别的直接方法。使用统计处理后的曲线比原观测值曲线更容易判别出地震前兆异常。在用统计方法识别出趋势性异常的基础上,可进一步采用模糊数学方法来识别中、短期地震前兆异常。 本文的研究结果表明,用统计-模糊识别方法可重点预测在离该井300公里以内的北东—南西方向上发生的M_s≥4.5地震。 本文最后探讨了用统计-模糊识别方法得出的中、短期地震前兆的某些特征,并作了简要的讨论。
The observed data of natural gas escaped from Tang-18 well for seven years in succession have been processed by using statistic and fuzzy recognition methods synthetically in this paper. The statistic method used here is the variance analysis and the fuzzy method is a direct one of fuzzy pattern recognition. For discriminating precursory anomalies, it is much easier to use the statistically processed curve than the use of original curves. When a trend anomaly has been discriminated by the statistic method, the fuzzy mathematics method can be further ado-pted to discriminate the medium-short-term precursory anomalies. The results show that the statistico-fuzzy recognition method can mainly be used to predict earthquakes with Ma≥4.5 in NE-SW direction within a 300km range around this well.Finally, this paper has explored some characteristics of the meolium-short-term precur-sors discriminated by the abovementioned method and given a brief discussion on them.
出处
《地震》
CSCD
北大核心
1989年第5期19-27,共9页
Earthquake
