The imaging quality of electrical resistivity tomography (ERT) crucially depends on the electrode array configuration. Although the symmetrical optimized ‘Compare R’ (CR) method improves computational efficiency, restricting the search to the symmetrical data set inherently limits the imaging accuracy. To address this limitation, this paper proposes a multi-step optimized CR method that progressively explores both symmetrical and asymmetrical arrays to extend the search space and further enhance imaging accuracy. Numerical experiments demonstrate that the multi-step optimized array yields the highest average relative model resolution (0.646) and structural similarity index measure (0.668), surpassing the symmetrical optimized array (0.615 and 0.630, respectively). Field experiments on pipeline detection confirm that the proposed array accurately identifies the location and geometry of underground anomalies and achieves superior imaging accuracy. Applications in karst cavity exploration further confirm that the proposed array effectively detects the deep karst caves and the bedrock interfaces, as validated by borehole drilling. Additionally, the detection performance of both optimized arrays is evaluated at different depths. The results indicate that the multi-step optimized array preserves anomaly geometry and resistivity more reliably at greater depths, attributed to the accumulation of asymmetrical data points in deep regions, which results in a significantly higher data density.
Qu et al. (Fri,) studied this question.