基于模糊模式识别的贴近度法判别煤田测井岩性

Closeness Degree Methods to Identify Logging Lithology in Coalfield Base on Fuzzy Pattern Recognition

选取影响岩性识别的7个主要参数,结合某地实际测井资料,建立基于模糊模式识别法的岩性识别标准模型。分别计算了各个待定样本与3个标准模型的格贴近度、海明贴近度、欧式贴近度、最大最小贴近度及算术平均最小贴近度。分析了不同贴近度方法求得的贴近度,结果表明5种方法所识别的结果都正确;通过分析对比5种方法所求得的最大贴近度曲线,可知海明贴近度方法所得结果较为稳定、准确,适合该次岩性识别实验;验证了应用模糊模式识别法进行测井岩性识别的可行性和科学性。

In order to better solve the problem of logging lithology identification, we use fuzzy pattern recognition method to identify lithology. Combined with actual logging data ot a well in the field, we select 7 main lithology identification parameters to establish the standard model of lithology identification based on the fuzzy pattern recognition method. Each sample＇s and the three standard model＇s closeness degree are calculated which contains lattice approximate degree, Hamming closeness degree, Euclid approach degree, maximum and minimum closeness and the minimum closeness degree. The closeness degrees based on the above different methods show that all the five identified results are correct. Through analysis and comparison the maximum closeness degree curve by five methods is obtained, we can see the result of Hamming closeness degree method is more stable, accurate and most suitable for lithology identification in this experiment. This study also verifies that the application of fuzzy pattern recognition method to logging lithology identification is practical and scientific.