【摘要】 采用沉积学、测井地质学、数学方法以及地震方法等对含煤地层及煤层组进行识别,总结出4种含煤地层及煤层组识别方法,并利用地震方法对煤层组进行了预测。结果表明:逻辑判别法和蜘蛛网图法仍然是较为有效的以定性为基础的半定量化方法,而聚类分析法和人工神经网络法则是成功且有效的定量分析方法,但不是通用的模型;对于勘探程度较低的海域区,要综合运用4种方法才能有效地预测含煤地层和识别煤层。地震反演技术与方法对于识别含煤段、判别含煤地层的分布是十分有效的,通过对琼东南盆地多个地震剖面的研究发现三瞬技术法对于识别较厚煤层交互层(复合层段)是可行的。通过对C-93-79和C-52-79剖面进行各类波阻抗反演,发现基于模型反演和神经网络反演的结果与含煤层段有良好的对应关系,是预测含煤层分布较有效的地震识别方法。更多还原

【Abstract】 Based on the sedimentology,logging,mathematics,and seismic prospecting and so on,the coal-bearing strata and coal seams was recognized,four methods were concluded and the coal seam was forecasted by seismic characteristics.The results show that the program of logical comparison and clustering methodology are effective semi-quantitative methods.Clustering methodology and spider diagram are effective equity methods but not be in common use.Only by synthetically utilizing these four methods the coal-bearing stratum and coal seams can be forecasted exactly.It is also effective to recognize the coal-bearing strata and coal seams by using the seismic inversion.By study on the seismic sections of Qiongdongnan Basin,the characteristics of lower instantaneous frequency and high amplitude of vibration in the Yacheng Formation,it is proved the methods of three-instantaneous profile technology is feasble.By constrained impedance inversion of seismic sections of C-93-79 and C-52-79,it is found that there is consistency between the inversions of model and neutral network and the strata with coal.So it is an effective method to predict the distribution of the coal seams and coal-bearing strata.更多还原