Lithology Identification by Support Vector Machine Using Well Logging Data

Zhaojie Zhang, Shi Fang, Wei Shen

2018

Abstract

The Jurassic formation in the Fengcheng area of Junggar Basin has complicated lithofacies because of its depositional environment supported the rapid deposition of sediment from a nearby provenance. The main lithofacies of this formation are mudstone, fine-grained sandstone, medium-grained sandstone and conglomeratic sandstone. Based on core and well logging data from the study area, this paper summarizes the characteristics of the rock and analyzes the logging response characteristics of the lithology. We use acoustic(AC), compensated neutron(CNL), density(DEN), gamma ray(GR) and resistivity(RT)logging data as training and test samples to establish a lithofacies recognition model by using a support vector machine(SVM). Additionally, we use a genetic algorithm to optimize the kernel parameter σ and penalty factor C. The results show that the model predicts that the overall coincidence rate is 85.1%, which is better than that predicted from a back-propagation(BP) neural network, and the model clearly improves the lithofacies recognition accuracy and efficiency.

Paper Citation

in Harvard Style

Zhang Z., Fang S. and Shen W. (2018). Lithology Identification by Support Vector Machine Using Well Logging Data.In Proceedings of the International Workshop on Environment and Geoscience - Volume 1: IWEG, ISBN 978-989-758-342-1, pages 400-405. DOI: 10.5220/0007431004000405

in Bibtex Style

@conference{iweg18,
author={Zhaojie Zhang and Shi Fang and Wei Shen},
title={Lithology Identification by Support Vector Machine Using Well Logging Data},
booktitle={Proceedings of the International Workshop on Environment and Geoscience - Volume 1: IWEG,},
year={2018},
pages={400-405},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0007431004000405},
isbn={978-989-758-342-1},
}

in EndNote Style

TY - CONF

JO - Proceedings of the International Workshop on Environment and Geoscience - Volume 1: IWEG,
TI - Lithology Identification by Support Vector Machine Using Well Logging Data
SN - 978-989-758-342-1
AU - Zhang Z.
AU - Fang S.
AU - Shen W.
PY - 2018
SP - 400
EP - 405
DO - 10.5220/0007431004000405