基于支持向量机的测井资料幅度异常验收方法研究

Abnormal amplitude checking method of log data based on SVM

当前的测井资料质量标准基于人工验收制定,其中包含大量的定性标准,如何依据定性标准进行测井资料质量自动验收,是目前面临的一个难题,将支持向量机方法应用于测井资料幅度异常预测,在资料中挑选可疑样本,依据已有的专家验收结论建立预测模型,对测井资料幅度异常进行预测,避免了自动验收中必须将定性标准定量化的问题,在实际应用中有较好的效果,对同类工程问题的解决有一定的借鉴意义。

The current log data quality standards are developed based on artificial checking. It contains many qualitative standards. How to check log data under the qualitative standards is a difficult problem currently. The SVM method is used in abnormal amplitude checking of log data, suspicious samples are selected from log data, prediction model was established on the basis of log data and its checking result was given by experts. It can predict abnormal amplitude checking result, which avoids the problem that qualitative standards must be inverted into quantitative standards. It has shown good effects in practical application, so it is worth generalizing.