Temporal and Spatial Data Fusion Integration in B-ANN for Improved Lithofacies Classification and Reservoir Characterization

Abstract

Objective: Using a Bayesian-Artificial Neural Network (B-ANN) model, the goal is to incorporate intricate previous geological knowledge to increase the precision and dependability of lithofacies classification in the characterization of hydrocarbon reservoirs.

Methods: Multiple geological data sources are integrated using a Bayesian Artificial Neural Network (B-ANN) model, which improves classification performance by fusing artificial neural networks with Bayesian inference. The classification accuracy of 97.6% is attained by the B-ANN methodology that has been suggested. Compared to current lithofacies classification techniques, this accuracy is noticeably better.

Novelty: Enhances performance of the B-ANN model by incorporating intricate past geological knowledge. optimally classifies geological data by integrating disparate sources of information seamlessly.