| 摘要 |
The invention described herein provides for an intelligent or "adaptable" filtering method that could be applied when filtering log data. In the present invention, for a given filter model, the adaptive filtering technique differentiates between true changes in the formation or borehole quantity measured and what can be considered to be just a "statistical fluctuation". So, when in long constant-value intervals (data deviations do not exceed those predicted by statistical theory) the filter will automatically increase the number of filter levels used, thereby providing a large reduction (smoothing) of the statistical variations. However, when approaching a spot where there is a true change in the quantity being measured (data deviations over and above those predicted by statistics) the filter will automatically and smoothly decrease, on both sides of this spot, thereby preserving the sharpness of the bed boundary or borehole fluid contact and also preserving the true shape of a thin geologic bed.
|
