PEMODELAN INVERSI DATA VERTICAL ELECTRICAL SOUNDING (VES) MENGGUNAKAN HYBRID DIFFERENSIAL EVOLUTION FLOWER POLLINATION ALGORITHM (HDEFPA)

Abstract

Geoelectric Resistivity is a geophysical method for mapping subsurface resistivity. Vertical Electrical Sounding (VES) is often used because of its ease in providing information on vertical lithological distribution. Modeling inversion of VES data is a complex challenge due to the non-linear relationship between data and model parameters. This research uses the Hybrid Differential Evolution Flower Pollination Algorithm (HDEFPA) as a global optimization method to overcome this problem. HDEFPA combines Flower Pollination (FPA) and Differential Evolution (DE) algorithms, which have both global and local search capabilities. Tests were carried out with synthetic VES data on two-layer, three-layer (K-type and H-type), and four-layer earth models. In addition, field data from seawater intrusion in Pelauw, Maluku, Indonesia were used. The results show that HDEFPA is superior to FPA and single DE in fast convergence and fewer iterations. The HDEFPA algorithm successfully validates field data, producing reliable and accurate results. This research provides insight into the advantages of global optimization methods over local inversion methods in inversion modeling of VES data.