![]() |
(tel) 617.253.8027 (fax) 617.253.6385 |
A Knowledge-Based Approach to Full Wave Data Processing
Marc Larrere
Submitted
to the Department of Earth, Atmospheric, and Planetary Sciences on January
16, 1987 in partial fulfillment of the requirements for the degree of Master
of Science in Geophysics
Abstract
A new approach to the processing of sequences of full waveform acoustic logs
is investigated. The rationale for this approach is primarily based on the
observation that processing and interpretation tasks strongly depend on each
other. Hence, a system that incorporates geologic knowledge in data processing
naturally and uses processing results for petrophysical evaluation can improve
the overall geological interpretation. The implementation of such ideas requires
the use of a versatile computer environment, allowing numeric and symbolic
processing. The new generation of Lisp machines satisfies these characteristics.
An interactive environment for the processing of sequences of acoustic signals
was designed using object-oriented programming, the package includes a novel
method for acoustic full waveform signal matching that uses dynamic time warping.
The system is tested on synthetic data, and field data are processed.
The feasibility of an expert consultant system for full waveform processing
and interpretation was investigated by implementing a prototype knowledge-based
system performing qualitative reasoning in the rock physics domain. Examples
of reasoning processes are presented and discussed.