Boosting Approach For Sonic Log Prediction Using Wavelet Transform

Sonic wave velocity is an important parameter in the oil and gas industry, as it provides crucial information about subsurface formations. The proposed method utilizes well logging data, specifically, the sonic log, which keeps track of how long a sound wave takes to navigate across the geological structures. Traditional methods for measuring sonic wave velocity involve drilling and core sample analysis, which can be time-consuming and expensive. Compressional transit time (DTC) and shear transit time (DTS) reports are not available for all wells drilled in the field due to resource constraints. In this situation, ML techniques can be used to predict compressional transit time and shear transit time data. Recently, non-invasive observations such as acoustic data from boreholes or seismic recordings have been used in conjunction with signal processing techniques to forecast sonic wave velocity. Wavelet Transformation, as one of the Signal Processing Techniques has been used to predict sonic logs, the process involves predicting using normal log data with the help of the XG boost model and then predicting it again once wavelet transformation has been done on the log data. Concerns regarding the accuracy and the amount of improvement were also addressed by comparing the two observations. The conclusion of this study can lead to sustainability in terms of sonic log prediction