Optimization of Equatorial Dipole-Dipole Antenna Geometry Using Evolutionary Computing for Subsurface Imaging

Abstract

Efficient mapping of buried utilities requires vital subsurface imaging. However, traditional galvanic dipole antennas suffer from current leakage due to suboptimal geometry and necessitate destructive borehole methods. To address this, an evolutionary computing approach optimizes a single-pair equatorial dipole-dipole antenna for high-frequency capacitive ground coupling at 38 MHz with 1.2 m skin depth. By minimizing the induction number (β) to 0 and maximizing the electrostatic geometric factor (K) to 1, quasi-static condition is achieved. Multigene genetic programming constructs fitness functions using antenna geometry parameters: dipole spacing, dipole length, and plate elevation. The optimization utilizes a genetic algorithm (GA) and evolutionary strategy (ES) to find the optimal parameter combination. GA enables faster exploration, while ES facilitates faster exploitation. Results indicate that an optimized equatorial antenna design requires plate elevation to be ≤ 2.24% of the combined dipole spacing and length. The ES-optimized antenna exhibits minimal power loss, high efficiency, isotropic gain, and direct radiation of the electric field on the pipe surface. This study eliminates trial and error in dipole design by presenting an effective technique for antenna optimization in subsurface imaging. The proposed approach achieves the global best solution, improving the cost-effectiveness and efficiency of subsurface mapping.