A Fully-Self-Adaptive Harmony Search GMDH-Type Neural Network Algorithm to Estimate Shear-Wave Velocity in Porous Media

Shear wave velocity (VS) is one of the most important parameters in deep and surface studies estimation geotechnical design parameters. This parameter widely utilized to determine permeability porosity, lithology, rock mechanical parameters, fracture assessment. However, measuring this either impossible or difficult due challenges related horizontal deviation wells difficulty reaching cores. Artificial Intelligence (AI) techniques, especially Machine Learning (ML), have emerged as efficient approaches for dealing with such challenges. Therefore, considering advantage ML, current research proposes a novel Fully-Self-Adaptive Harmony Search—Group Method Data Handling (GMDH)-type neural network, named FSHS-GMDH, estimate VS parameter. In way, Memory Consideration Rate (HMCR) Pitch Adjustment (PAR) are calculated automatically. A method also introduced adjust value Bandwidth (BW) based on cosine each decision variable values. addition, variable-size harmony memory proposed enhance both diversification intensification. Our FSHS-GMDH algorithm quickly explores problem space exploits best regions at late iterations. allows training prediction model P-wave (VP) bulk density (RHOB). Applying carbonate petroleum reservoir Persian Gulf demonstrates that it capable accurately estimating better than state-of-the-art machine learning methods terms coefficient determination (R2), Mean Square Error (MSE), Root (RMSE).