Multiobjective Optimization of Mobile Coplanar Array Capacitance Sensor for Imaging Internal Damages in Asphalt Mixtures

Coplanar capacitance imaging technology (CCIT) is advanced in detecting and identifying internal damages in measured objects. However, typical coplanar array capacitance sensors (CACSs) exhibit contradictory performances in terms of signal strength, imaging accuracy, and detected depth. Currently, there is few CACS capable of identifying internal damage in asphalt layers accurately. To address this gap, a novel mobile CACS (MCACS) is proposed and optimized using response surface methodology (RSM) combined with a nondominated sorting genetic algorithm III optimization (NSGA-III) algorithm to image internal damages in asphalt mixtures. Initially, a central-composite design (CCD) of RSM is used to develop the statistical models of the three optimized performances of MCACS. The significance of structure parameters on the three sensor performances is then analyzed using ANOVA and 3-D response surface diagrams. Furthermore, the NSGA-III algorithm is used to optimize the multiperformances of MCACS. In addition, the optimized performances of the MCACS are validated using laboratory experiments. All three statistical models are determined to be significant since the R-squared is greater than 0.95. All of the single structure parameters are significant for the three sensor performances. Based on the optimization of the NSGA-III algorithm, the optimal solutions are predicted and confirmed with less than 5% error from the simulation results. Compared to the traditional CACS, the MCACS can identify internal damage in asphalt mixtures with optimal signal strength, uniformity of sensitivity field distribution, and penetration depth more accurately.