Abstract

Coplanar capacitance detection exhibits remarkable advantages in the detection of road texture in asphalt layers, including high sensitivity and minimal environmental constraints. However, the inherent performance contradiction between signal strength and penetration depth of traditional interdigitated coplanar capacitance sensors (ICCSs) has restricted their widespread application in road texture detection. To address this issue, a hybrid approach combining response surface methodology (RSM) and non-dominated sorting genetic algorithm II (NSGA-II) is developed to optimize the structural parameters that influence the signal strength and penetration depth of a novel ICCS. Initially, a central-composite design (CCD) based on RSM is employed to establish statistical models for the two key sensing performances of ICCSs, namely signal strength and penetration depth. Subsequently, Analysis of Variance (ANOVA) and three-dimensional (3D) response surface plots are utilized to investigate the significant effects of various structural parameters (electrode length, width, and inter-finger gap) on the two sensing performances. Furthermore, NSGA-II is applied to search for global optimal solutions using the established statistical models, thereby achieving multi-performance optimization of the ICCS. Finally, the fabricated ICCS is used to detect the surface texture of asphalt mixture specimens with different gradations, and the results are compared with those obtained by laser point cloud detection. The results indicate that both statistical models are highly significant, with the coefficient of determination (R-squared) exceeding 0.95. All individual structural parameters have a significant impact on the two sensing performances. Based on the optimization by the RSM-NSGA-II hybrid method, the predicted optimal parameters are verified, showing a relative error of less than 5% from the simulation results. Additionally, the detection results of the ICCS are consistent with the laser point-cloud data, demonstrating its feasibility for pavement texture detection.

IPC Classification

Keywords